June 2021 - ACS Axial | ACS Publications
Search
Close

Announcing the Journal of Medicinal Chemistry Early Career Advisory Board

Journal of Medicinal Chemistry (JMC) is excited to announce the appointment of its Early Career Advisory Board. The new Early Career Advisory board features over a dozen young researchers with a variety of expertise from across the globe. They will work with journal Editor-in-Chief Professor Craig Lindsley and its Associate Editors by sharing their perspectives and ideas on evolving topics in the medicinal chemistry community.

Yan Li, Department of Medicinal Chemistry, School of Pharmacy, Fudan University

Personally, becoming a member of the JMC Early Career Board will allow me to better follow the hot topics in the field of medicinal chemistry, and provide the opportunity to meet and communicate with the distinguished young medicinal chemists from all over the world, which is very beneficial to my scientific research career. At the other hand, I hope to promote the outstanding research work of Chinese scientists and help to enhance the academic influence of Chinese scientists worldwide. Hope to make my modest effort in the future work.

Paul C. Trippier, Director, Pharmaceutical Sciences Graduate Program, University of Nebraska Medical Center

The Journal of Medicinal Chemistry is the most prestigious in our field. I am excited to play a small part in continuing this tradition and steering the journal forward. I see serving on the Early Career Board as an opportunity to reach out to new talent in the field, bring them into the collection of amazing scientists who publish in the journal and to increase collaborations with the ACS MEDI Division.

Fleur M. Ferguson, William A. Lee Assistant Professor, Department of Chemistry & Biochemistry

I’m incredibly excited to join the JMC Early Career Board! I am looking forward to learning more about the publishing and editorial process, and collaborating with an outstanding international team of early-career scientists and the editorial board to shape the future of JMC.

I hope to contribute by working with the Early Career Board to identify and highlight important research advances and emerging themes, particularly those at the interface of chemical biology and medicinal chemistry. In addition, I hope to bring an early-career perspective to journal policy issues.

Deepak B. Salunke, Assistant Professor and Ramalingaswami Fellow, India Department of Biotechnology, Coordinator, National Interdisciplinary Centre of Vaccine, Immunotherapeutics and Antimicrobials, Department of Chemistry, Panjab University, Chandigarh

Being a Journal of Medicinal Chemistry early career board member is an exciting opportunity. A journey with diverse group of young minds with an experienced and qualified leadership will definitely be a super learning experience. While working for the journal to recruit high-quality submissions and carry out other related activities, I am very much interested to use this opportunity to build a discussion/interaction platform of expert medicinal chemists at India and compile yearly virtual issues showcasing the medicinal chemistry happening in India at both academia as well as at the industrial settings. There are many good organic chemists in academia and their transformation via building good discovery working groups and medicinal chemistry platforms is definitely required. Development of a new trend medicinal chemistry curriculum for master students at the recently initiated National Interdisciplinary Centre of Vaccine, Immunotherapeutics and Antimicrobials at the Panjab University, Chandigarh is another short-term goal. This association with the JMC and other expert board members will be very useful and crucial towards this direction. I also would like to take this opportunity and coordinate yearly international meetings where experts from all over the world can showcase the research activities at the interface of chemistry and biology to educate the young minds towards the exciting field of Medicinal Chemistry.

Amanda L. Garner, Associate Professor, College of Pharmacy, Department of Medicinal Chemistry, University of Michigan

I am excited to be a part of the JMC Early Career Board to be able to provide insights into the future of the journal. My research spans chemical biology and medicinal chemistry, and I am looking forward to serving as a bridge between these fields to expand the content and readership of JMC. Moreover, as a woman in medicinal chemistry, I am especially eager to provide new ways to highlight the important contributions of women to the field to inspire our next generation of female medicinal chemists.

Caitlin Kent, Scientist, Sanofi

I’m honored and excited to be part of such an impressive group of young medicinal chemists and represent industry within the Early Career Board. I hope my participation within the Early Career Board and other journal matters will ultimately influence the content and medicinal chemistry stories that train future generations of medicinal chemists.

Pedro M. Garcia Barrantes, Research Scientist II, Vertex Pharmaceuticals

I am honored to be part of the Early Career Board. JMC is a unique journal, the staple publication in the field, and I am thrilled to be part of the team that makes it happen. This board is a great initiative and I am impressed by the enthusiasm and the diversity of backgrounds of the selected scientists, I believe everyone bring something especial to the board. I look forward to collaborating with other members to provide the perspective of a younger scientist audience.

I believe the journal plays a significant role in educating the next generation of medicinal chemists. I have been reading JMC since before I decided to be a med chemist, and it has had an enormous impact on my training. I would like to contribute to keep the high scientific quality of the journal, bringing topics that are fresh, current and of interest to the drug discovery community, and by attracting and facilitating the publication of ground-breaking discoveries. I believe that if we work hard on this, we will continue to have an impact in the field and towards the drug discovery’s endgame, the development of new treatments for patients in need.

Benita Sjogren, Assistant Professor of Medicinal Chemistry and Molecular Pharmacology, Purdue University

It is an honor to join the JMC Early Career Board. As a member I hope to help narrow the gap between biologists and medicinal chemists in early drug discovery. I look forward to working with the members of the JMC Early Career Board to increase visibility for the journal, and help guide where it will go in the future.

Marie-Gabrielle Braun, Genentech, Inc.

I am very excited to be part of the JMC Early Career Board as JMC is the most trusted and cited medicinal chemistry journal in the world. I hope to bring my industrial experience to contribute to the journal’s visibility by further expanding its scope, authorship, and impact. I want JMC to be the first choice for early career researchers to publish their medicinal chemistry research so that their research gets the most visibility across the community. In addition, I hope to address gender and ethnicity parity issues by supporting women and minorities who work in medicinal chemistry.

Eufrânio N. da Silva Júnior, Professor of Organic Chemistry, Institute of Exact Sciences, Department of Chemistry

JMC is a highly appreciated and respected journal by the global scientific community engaged in Medicinal Chemistry. Contributing to the JMC is an aspiration present in the minds of researchers working in this field of science. I have experience working in Medicinal Chemistry since the beginning of my career and obviously, I share this feeling. Sailing in uncharted waters towards the discovery of new bioactive molecules demands a complex effort, collaborative work, and a mutual desire to reach the state-of-the-art in frontier topics that are often initially inaccessible and are achieved with dedication and resilience. JMC is the materialization of the purpose attained, and being part of this process is both hugely motivating and exciting. I hope to continue contributing to the continued building of a community of Medicinal Chemistry researchers related to the JMC and to highlight the work of early career investigators and researchers from countries that face significant structural challenges to carry out projects at the frontier of science, and who persevere with dedication and love for science. I would also like to highlight the power of Medicinal Chemistry for all scientists who work in a multidisciplinary approach to achieve unbelievable achievements. Finally, I hope to improve my knowledge, work collaboratively, and build networks of friendship and collaboration around the topic of Medicinal Chemistry.

Christian W. Gruber, Associate Professor, Drug Discovery and Peptidomics, Medical University of Vienna Center for Physiology and Pharmacology

I am excited to serve on the JMC Early Career Board to contribute to the publication of high quality manuscripts in the top journal of its field. During my term I would like to help shape the future of Medicinal Chemistry, by suggesting themed issues or perspectives, and by building links between young scientists and the leaders in the field.

Haruto Kurata, Senior Scientist, Medicinal Chemistry Research Laboratories

I am so excited about interacting with various scientists all over the world through my role as a member of the JMC Early Career Board. I would like to influence early career medicinal chemists through my expertise in the field, which I hope leads to promoting JMC. I also plan to contribute to promoting cultural diversity, which I believe is very important to enhancing science.

Carles Galdeano, Serra Hunter Lecturer Professor, University of Barcelona

Since the beginning of my research career in drug discovery and medicinal chemistry I have considered JMC the reference journal in the field. The opportunity to contribute to build the reputation and keep the excellence of the journal as a member of the Early Career Board is very thrilling. The JMC Early Career Board is made up of young but top-level scientist from around the world. I hope that we can work together, in a synergistic manner, to help high-quality and innovative science to be publish efficiently in JMC. In parallel, I would like to help to increase the visibility of early career researchers in our field promoting different initiatives.

I was thrilled at being named a part of the JMC Early Career Board. To me, the JMC is not only the most trusted and cited medicinal chemistry journal in the world, but also a valued source where to find the latest cutting-edge advances in drug discovery and useful content that supports my day-to day training as medicinal chemist. I owe the JMC a great deal for tremendously developing the field over the years and aiding me in getting better at my job.

As a member of the JMC Early Career Board, I am eager to bring to this role both industrial and academic experience and positively contribute to the growth of this prestigious journal. I am committed to retaining and expanding the legacy of JMC, ensuring a fair and transparent publishing process, as well as supporting the journal to broaden and diversify its scope, in terms of scientific content and key players, such as readers, authors, and reviewers, to strengthen the research community.

Federica Prati, Medicinal Chemistry Scientist, Presso Angelini Pharma

I was thrilled at being named a part of the JMC Early Career Board. To me, the JMC is not only the most trusted and cited medicinal chemistry journal in the world, but also a valued source where to find the latest cutting-edge advances in drug discovery and useful content that supports my day-to day training as medicinal chemist. I owe the JMC a great deal for tremendously developing the field over the years and aiding me in getting better at my job.

As a member of the JMC Early Career Board, I am eager to bring to this role both industrial and academic experience and positively contribute to the growth of this prestigious journal. I am committed to retaining and expanding the legacy of JMC, ensuring a fair and transparent publishing process, as well as supporting the journal to broaden and diversify its scope, in terms of scientific content and key players, such as readers, authors, and reviewers, to strengthen the research community.

Mariateresa Giustiniano, Assistant Professor, University of Naples Federico II | UNINA · Department of Pharmacy

I am very grateful and excited to serve as a member of the Early Career Board of JMC, the leading journal in disseminating the most important advances in drug discovery! I am looking forward to working with my colleagues to assist the Editorial Board at my best and to share ideas aimed at promoting the training and the interests of the younger generations of medicinal chemists.

Yimon Aye, Swiss Federal Institute of Technology Lausanne (EPFL)

Image Credit: Jon Reis

My laboratory develops original chemical biology tools with which to solve challenging research problems in the broader fields of biological stress regulation and genome maintenance. Both areas are a treasure trove offering untapped opportunities for drug discovery, in alignment with the spirit of medicinal chemistry research. As a scientist trained in both organic chemistry and classical enzymology, I am enthusiastic to help promote Journal’s ongoing and future efforts, particularly those involving interdisciplinary knowhow.

Damian W. Young, Assistant Professor, Baylor College of Medicine, Depart of Pharmacology and Chemical Biology

I am honored to be a member of the JMC Early Career Board. JMC was actually the first scientific journal that I was exposed to as an undergraduate student during my first summer internship at GD Searle and Company. As I read articles, I can still recall how excited I was to connect the new passion that I had found for synthetic organic chemistry with molecules that possessed biological activity. Fast forward to the present, I am excited to be a part of the Early Career Board and to provide my insight about future directions of this important journal. Given that I work within a medical college instead of a traditional chemistry department I hope to provide views for the journal that are relevant to translation.

Get to Know: Kwabena Bediako

Kwabena Bediako is an Assistant Professor in the Department of Chemistry of the University of California, Berkeley and a member of the Editorial Advisory Board for the Journal of the American Chemical Society (JACS). Read on to learn more about his life and work.

Describe the current focus of your work.

My group works to find more energy efficient ways to direct charge transport within solids and at solid–liquid interfaces. We look to tune these processes using materials that are only a few atoms thick, so-called “two-dimensional” or “2D” materials. Graphene is the prototypical 2D material, but this is a very large and diverse family of materials with unique degrees of freedom for manipulating physical and chemical behavior. Ultimately, we hope that the phenomena and fundamental principles we discover can serve as the basis for new ultralow power electronic devices and highly efficient electrochemical energy conversion systems.

How did you become interested in your field?

I have always been interested in scientific problems related to energy sustainability (that is, after I got over my childhood dream of becoming a pilot). I read articles about the potential of solar energy to solve some of the critical energy and environmental challenges we face as a global community. Growing up in Ghana—a country that, like many others in Africa, is blessed with a lot of sun—I was eager to do research in a field that I felt could make this a reality. My undergraduate studies introduced me to inorganic chemistry and as a graduate student I studied electrochemical water splitting (with a view to store renewable electricity in fuels). My postdoctoral research in a condensed matter physics lab helped me to appreciate the outstanding challenges for charge transport in solids and the energy currently wasted in conventional electronic devices. My research program now tries to find answers to questions at the nexus of these fields.

What does becoming a part of the JACS Editorial Advisory Board mean to you? What are your hopes for the journal?

The invitation came as quite a surprise and it is a great honor to be a part of the EAB of JACS. For me, JACS has always represented a forum at the pinnacle of scientific thought in chemistry. I know this will continue long into the future. I echo Editor-in-Chief Erick Carreira’s vision for the journal to continue to serve as a flagship platform for both fundamental discoveries and interdisciplinary advances in the chemical sciences.

Black chemists continue to be underrepresented. What changes need to take place in the chemistry field to change this?

That is quite a complex problem and I am far from an expert on that subject, but clearly there are institutional issues to overcome at nearly all levels in terms of both recruiting and retention. It is well documented that one of the root challenges to increasing the proportion of racial and ethnic minorities in STEM fields in the U.S. is the creation of a sense of belonging. So, any solution to this challenge must include a concerted effort to foster a community of these researchers in which common experiences can be shared, along with efforts to increase the visibility of the scientific research and accomplishments of these groups. Admittedly, some of the issues have become so ingrained that they will take time to be overcome, and not everyone is necessarily in a position to directly bring about some of the biggest policy changes that are needed. Still, I think what we can each do as individuals is to learn about, confront, and work to overcome our implicit biases. Everyone has that personal agency. I think that is the challenge to each of us, and a challenge we can begin addressing on our own, right away to chip away at one part the problem.

What’s one piece of advice you wished you’d received before starting your career in chem?

Hmm… maybe, “take a management course.”

Where do you hope to see the field, as it pertains to Black scientists in the next 10 years?

I hope that the proportion of Black scientists at all levels at least reflects the population demographics of the country.

What chemist has inspired you most?

Alice Ball’s story is quite inspiring. She had such an impact on her time in an unfortunately short life.

Read Kwabena Bediako’s Research in ACS Publications Journals

ACS Publications Begins Accepting Ombudsperson Applications

The search for ACS Publications’ first-ever ombudsperson is underway with the publication of a call for proposals for the role. This call comes on the heels of a June 2020 special joint editorial outlining five steps our organization would take to address systemic racism in our publishing processes:

  • Gathering and making public our baseline statistics on diversity within our journals, encompassing our Editors, advisors, reviewers, and authors; annually reporting on progress.
  • Training new and existing Editors to recognize and interrupt bias in peer review
  • Including the diversity of journal contributors as an explicit measurement of Editor-in-Chief performance
  • Appointing an ombudsperson to serve as a liaison between Editors and our Community
  • Developing an actionable diversity plan for each ACS journal

These steps are highly interdependent, and ACS Publications has made steady progress over the last year, working on these goals in tandem.

One of the most challenging steps was the move to create a new ombudsperson role at ACS Publications, a position unlike any the organization has had before. This role will act as an independent liaison between ACS Publications and the chemistry community to address concerns about Editors, Editorial Advisory Board members, or reviewers regarding editorial processes. The ombudsperson will help ACS Publications achieve our goal of providing peer review services that are as inclusive as possible to the chemistry community, regardless of gender, sexual orientation/identity, race/ethnicity, religion, employment status, or status in the field.

The request for proposals was developed after extensive work with the International Ombudsman Association to learn more about the function of an ombudsperson role and how best to implement and support such a position. ACS Publications is currently accepting applications and hopes to have someone in the role this fall.

Candidates should have the following qualifications:

  • Minimum of at least 3 years of experience as an organizational Ombuds or graduate education in conflict resolution, law, or a related field; and ombuds training (completion of International Ombudsman Association Foundations or equivalent).
  • Must be familiar with the IOA Code of Ethics and Standards of Practice. Being a member of the International Ombudsman Association is not necessary but preferred.

A scientific background is desirable but not required.

Application Process:

Applicants for this position should apply on or before 5 P.M. EDT on August 13, 2021.

Proposals should include the following: A CV and 2 letters of reference from experienced organizational ombuds practitioners, and a review of ACS Publications policies, and compensation expectations.

Applications should be sent to: Ombuds_Applications@acs.org

Read 2021 Editorials on Racial Issues in Chemistry from ACS Publications Journals:

Queer, PoC, Creative, STEM
Anal. Chem. 2021, 93, 21, 7541–7542
DOI: 10.1021/acs.analchem.1c01826
***
Introducing Analytical Chemistry’s Diversity and Inclusion Cover Art Series
Anal. Chem. 2021, 93, 3, 1211–1212
DOI: 10.1021/acs.analchem.0c05466
***
Shaping the Future of Higher Education: Practical, Community-Driven Initiatives to Improve Academic Climate
ACS Cent. Sci. 2021, XXXX, XXX, XXX-XXX
DOI: 10.1021/acscentsci.1c00491
***
A Reflection on Juneteenth and the Diversity of Our Chemical Neuroscience Community
ACS Chem. Neurosci. 2021, XXXX, XXX, XXX-XXX
DOI: 10.1021/acschemneuro.1c00378
***
Black Scientists Are Not the Door to Diversity
ACS Chem. Neurosci. 2021, XXXX, XXX, XXX-XXX
DOI: 10.1021/acschemneuro.1c00375

Call for Papers: From Reaction Informatics to Chemical Space

The Journal of Chemical Information and Modeling is currently preparing a special issue focusing on recent developments in modeling the synthetically accessible space of small organic compounds. Modeling chemical space is critically important in enabling machine learning to drive pharmaceutical and biotechnological research. This special issue aims to give a broad overview of this emerging, exciting technology and will provide visibility for authors’ excellent science.

Read the Editorial

Submission Instructions

Manuscripts must adhere to the guidelines available on the Journal of Chemical Information and Modeling page and must be submitted electronically through the ACS Paragon Plus portal.

In Paragon Plus, specify a manuscript type, and activate the special issue feature to designate the paper for ‘From Reaction Informatics to Chemical Space.’ In addition, state in your cover letter that the paper is being submitted for the special issue.

All invited and contributed manuscripts will be screened for suitability upon submission and undergo the standard peer-review procedure of the journal. The final submission deadline for inclusion in the special issue is September 1, 2021.

ACS Publications’ Name Change Policy Advances Inclusion in Scholarly Publishing

At ACS Publications, we’re committed to creating an inclusive experience for all of our author, reader, and reviewer communities by championing an environment that advances and celebrates diversity, inclusion, and respect. As we celebrate Pride Month 2021, we want to share the changes that ACS is undertaking to create a more inclusive experience for our authors, and reflect on how our initiatives are impacting the entire sector of scholarly communications.

In September of 2020, ACS announced a new trans-inclusive policy to allow authors to change the names used on their previously published articles. This policy was designed with the needs of the transgender scientific community in mind, and with the generous assistance of several members of that community. In the nine months, the policy has been in place, we have already received requests to update over 200 publications from more than 30 authors, with some articles dating back several decades.

Under ACS’ policy, an author who requests a name change for any reason is treated with confidentiality and respect. The author is not asked to provide proof or documentation of their name change, and a name change is not treated as a correction to their paper. Thus, any co-authors are not alerted to the change and no public notice is added to the paper. The ACS policy also ensures that all other references to the author’s identity, including pronouns, salutations, captions, and other elements of the paper, are updated appropriately. ACS was the first chemistry publisher to adopt this policy and shared its lessons learned from implementation with other publishers to encourage them to adopt similar policies.

Following ACS Publications’ lead, other scholarly publishers have announced similar policies or have publicly confirmed that they’re working on similar policies. The impact has gone far beyond the field of chemistry, as two working groups from the National Information Standards Organization and the Committee on Publication Ethics have formed to address similar issues.

ACS recently joined the Coalition for Diversity & Inclusion in Scholarly Communications (C4DISC) as a gold partner. C4DISC is jointly hosted by the Society for Scholarly Publishing (SSP) and Association of University Presses (AUP) and was formed to provide a venue to discuss and address issues of diversity, equity, and inclusion within the scholarly communications industry. C4DISC has an active working group focusing on inclusive language and image guidelines that we plan to be involved with, as we work to create a more inclusive community across publishing.

ACS Publications continues to work toward a more positive and gender-inclusive experience for authors at all stages of their publication journey. We are presently in the process of removing gendered honorifics from ACS Paragon Plus account profiles and gendered pronouns in journal correspondence, replacing them with gender-neutral language. New authors and reviewers have the option to choose from an updated list of gender-neutral titles.

How can ACS Publications continue to improve inclusivity in scientific publishing? Share your thoughts via our bias feedback form or in the comments below.

Machine Learning in Chemistry: Now and in the Future

ACS In Focus recently held a virtual event on “Machine Learning in Chemistry: Now and in the Future” with Jon Paul Janet, Senior Scientist at AstraZeneca and co-author of the ACS In Focus Machine Learning in Chemistry e-book.

This event had a brief discussion of Dr. Janet’s ACS In Focus e-book, a conversation on the future of machine learning, and a presentation on the exciting research Dr. Janet and his colleagues have recently done using machine learning to accelerate the search for new materials.

Below you can watch the recording of the webinar and view some questions your colleagues asked.

View the Webinar Recording:

Interested in learning how to get access to Machine Learning in Chemistry? Talk to your librarian today!

Read Dr. Janet’s Answers to Community Questions

As a beginner without prior knowledge of programming language, how can one go into this field? What are the prerequisites one needs to acquire for this field?

I think learning some basic python scripting is the best way to get started, because there is a great community and tons of tools that can help make trying machine learning on chemical problems easy – sklearn and RDKit are amazing and get you quite far. But these need some familiarity with scripting at least.

How do you compare machine learning with quantitative structure-activity relationship (QSAR) modeling, which has been around for 30 years?

I would say “QSAR modeling” is a label for a specific type of machine learning (activity prediction based on molecular structure). Many QSAR techniques such as SVR/SVM and random forest are components of traditional/shallow machine learning, so we have been doing “machine learning in chemistry” for decades and in my experience, traditional QSAR methods are mostly competitive with deep learning approaches for affinity prediction especially. But I don’t think that is the whole story – new machine learning methods are letting us solve new types of problems such as generative models, retrosynthesis prediction, massive multitask predictions, etc. that don’t fit neatly into the QSAR label. In these cases, they are not really competing with other methods as much as expanding the type of data we can use and using it in new ways. If deep learning methods beat canonical QSAR approaches depends on who you ask, but in my experience, one is almost never worse off with ChemProp instead of a fingerprint method (though one might not be as much better off as one hopes). Neural networks also let us do interesting things in QSAR space such as multitask learning or even federated learning, and I think these approaches will be the standard in the future.

Can we simulate the effect of magnetized water clusterization through machine learning tools?

I don’t know much about water magnetization, but I know there has been some work in simulating the behavior of water clusters with neural networks. I am not sure if anyone has used these methods to infer magnetic properties though!

What are the job prospects for someone with a background in chemistry and computer science if he/she wants expertise in machine learning in the context of computer science?

Good question. There is definitely interest in industry, both in pharmaceuticals but also increasingly in materials design, including lots of startups in the last few years. I think there are roles for people with both more computer science experience and more chemistry experience. I think ideally a team would be comprised of people with both backgrounds.

What would be more appropriate machine learning techniques in de novo drug discovery?

This is what I work on now and it is an open question. I think SMILES-based generative models seem to be performant enough that more complex graph-generating methods don’t seem to pay off. I think the biggest issue is honestly getting useful scoring functions, so incorporating more multitask QSAR and physics-based/assisted approaches.

Do you envision a phase of disappointment after the current hype of machine learning, similar to what has happened with other promising technologies?

You can read about the “AI winter” of the 1980s. If you haven’t: machine learning has had this hype-crash cycle before. I think the current hype around general machine learning comes from a few high-profile technologies of the last decade-ish: convolutional neural nets for images, recurrent networks (and now transformers) for text, and advances in reinforcement learning for game playing. Each of these has followed in rapid progression and made a big difference in their respective domains and helped maintain and build hype. Without another flashy advance, I think we will see another slowdown. I am not sure we will see such a large ‘crash’ per se but I foresee more of a quieting down of interest, in some ways that might already be happening with, for example, generative models.

Does machine learning have applications in the design-of-experiment optimization process?

I would say machine learning is a big umbrella and yes, but I don’t put general optimization under machine learning. Instead, many of the DOE methods, at least the ones I use, depend on an explicit surrogate to predict how the objective function behaves on unseen points. Training of these surrogates, which are usually Gaussian processes or related, is a machine learning task. Maybe the DOE people would say the algorithmic choice of how you use this information is something else (i.e., not machine learning), but I don’t think the label really matters.

How could machine learning help ab initio algorithm minimize calculation times?

A whole lot of ways! One idea is to build a neural network to predict density functional theory (DFT) energies as a function of the structure, then you can run the simulation on a neural network potential, and only call the DFT to check on, and update, the potential as needed. Other clever people have used machine learning directly integrated into the hamiltonian, or to predict which orbital pairs to keep in a post-HF method. You can also accelerate geometry optimizations by bootstrapping a surrogate model at each step. In our work, we used machine learning models to construct good-quality starting geometries for our calculations, which reduces the number of optimization steps needed. We have also looked into using machine learning to predict when multireference methods are needed (vs. single-determinant DFT), which can help save a lot of time!

Do you have any comments or suggestions for the prediction of biological activity data? How do the new methodologies perform vs. QSAR models?

So we have been doing “machine learning in chemistry” for decades, the only difference now is that we have a larger toolbox of models that might or might not help build better activity models. QSAR methods in particular have benefited from a lot of optimization and seem to extract almost all the useful predictive power out of affinity data. If deep learning methods beat canonical QSAR approaches depends on who you ask, but in my experience, one is almost never worse off with ChemProp instead of a fingerprint method (though one might not be as much better off as one hopes). Neural networks also let us do interesting things in QSAR space such as large-scale multitask learning or even federated learning, and I think these approaches will be the standard in the future. These methods give us a way to overcome the typically limited amount of affinity data we have for a particular target by bringing in more information.

How accurate can the latest machine learning methods be in predicting the possible starting synthons for a designed novel molecule or polymer material?

I can only really comment that for drug-sized organic molecules we can actually make reasonable synthon predictions in many cases, e.g. machine learningPDS tools are used in the real world every day. As for polymers, I think there is a lot less published so I am not sure. In principle, yes, but it will depend on the data that is available. In my experience, these methods are extremely reliable at finding feasible, commercially available synthons for common disconnects (amide bonds, etc) but a little less reliable for more exotic chemistry. This is pretty cool because it shifts some of the busy work of synthetic chemists to focus on more interesting problems!

I am an organic chemistry Ph.D. student. I have started learning the basics of machine learning. Is it possible to work in the machine learning chemistry field in postdoctoral research, or is it already too late?

Definitely not! My Ph.D. group worked with a number of postdocs from purely chemistry backgrounds and there is a lot of domain experience that you gain from a Ph.D. that can be useful in applying machine learning methods. Probably I would recommend trying to join a group that does machine learning so you can learn from them. That said, being comfortable with python scripting (or some similar language) is pretty crucial and those skills take time to practice, so that might be a great additional skill to obtain. There are a lot of good online courses.

Have you ever tried to modify PTFE to design an inorganic-organic hybrid complex

No I haven’t, but it sounds interesting. I am more in the machine learning/comp chem side so I don’t know how easy it would be to do in practice.

How well is machine learning in predicting biological activity, and what does this prediction depend mainly on?

So we have been doing “machine learning in chemistry” for decades, the only difference now is that we have a larger toolbox of models that might or might not help build better activity models. QSAR methods in particular have benefited from a lot of optimization and seem to extract almost all the useful predictive power out of affinity data. If deep learning methods beat canonical QSAR approaches depends on who you ask, but in my experience, one is almost never worse off with ChemProp instead of a fingerprint method (though one might not be as much better off as one hopes). Sadly I don’t think we have gotten much better at activity prediction in the last few years, but neural networks also let us do interesting things in QSAR space such as large-scale multitask learning or even federated learning, and I think these approaches will be the standard in future. These methods give us a way to overcome the typically limited amount of affinity data we have for a particular target by bringing in more information. Some other limiting factors apart from dataset size are the quality of the data and the sensitivity to small structural changes (activity cliffs). These are pretty difficult to deal with since all machine learning models are smooth functions and struggle to learn large jumps (as humans sometimes do trying to rationalize SAR).

How many DFT calculations (training points) do you need to parametrize a reliable NN model?

It depends on how dissimilar the new structures you want to predict are from your training data. I would say a few hundred at least to predict static properties, though if you want a machine-learned potential (i.e., a force field) to predict dynamics you might need millions. One nice thing is as long as you keep an eye on uncertainty, you can be selective in which new data points you acquire.

I'm confused about saying you can't do a DFT model to compare on a large number of compounds. There are pretty good and cheap models such as cepC.

Is that a semiempirical method? The main issue is that the systems I was studying are open shells, and generally transition metals are not well-parameterized by these methods which don’t handle spin state ordering. This has implications for bond lengths and redox potentials. Another point is that machine learning methods trained on DFT data can often outcompete semiempirical methods while being much cheaper (once the DFT is done, of course!).

How many DFT calculations would be required to train the neural networks to suggest the next calculation?

In this work, I was doing about 100 DFT calculations each time, but it would be possible to do more or less. From a design of experiments perspective, it is actually optimal to update the model after every calculation, but this is time-consuming and inconvenient so some degree of batching is needed

Is it possible to use machine learning in environmental remediation? For example, in degradation of pollutants using photocatalysts?

Yes, I can’t see why not. machine learning is fairly general, and predicting some property of a certain molecule doesn’t really depend on what that property is (though different methods might be more or less suited to certain tasks). I have seen some work about predicting light-harvesting abilities but I don’t really follow remediation literature.

How could you augment the amount of data you obtain from the DFT so as to train your model?

There are a few neat ideas that can help. One idea I particularly like for chemistry is transfer learning, which is building a model to predict some kind of chemical property from a big database, and then fine-tuning it on some smaller dataset. For example, these authors trained a model on a large number of cheap(er)-to-compute DFT energies and then fine-tuned the model on more expensive coupled-cluster calculations. One can also do ‘unsupervised’ learning, where you train a model to predict some basic thing about the molecule such as hiding one atom or bond and training the model to supply the missing atom or bond label, as in this IBM work. In either case, you hope to teach the model the basic rules of chemistry, and it might be easier to teach it about some specific property with fewer examples (something like it is easier to understand polar molecules are good if you understand polarity). Another idea is data augmentation, which is done for image-based models by rotating or zooming in or out on the images, increasing the number of samples. This can also be applied to chemistry, for example, by writing multiple identical representations of the molecules. Neither case is as good as getting more data but sometimes that is all one can do.

Can we also use machine learning in the band-gap engineering of materials?

This one I can be more confident about: definitely! There are numerous examples. It seems like band gaps are quite easy to predict in many cases.

I find that with machine learning properties using quantum chemistry, the big issue is the lack of data. What are your thoughts on whether you have “enough” data - not just in amount but in variety, especially if using DFT data as opposed to experimental?

One nice thing about using QM data is that we can be really optimal in the choice of what data we acquire (as opposed to experiments where some materials would be very informative but we can’t make them). Picking data points to cover model blind spots can be very effective. Of course, there are still limits, especially for larger or complex molecules. One idea I particularly like for chemistry is transfer learning, which is building a model to predict some kind of chemical property from a big database, and then fine-tuning it on some smaller dataset. For example, these authors trained a model on a large number of cheap(er) DFT energies and then fine-tuned the model on more expensive coupled-cluster calculations. One can also do ‘unsupervised’ learning, where you train a model to predict some basic thing about the molecule such as hiding one atom or bond and training the model to supply the missing atom or bond label, as in this IBM work. In either case, you hope to teach the model the basic rules of chemistry and it is easier to teach it about some specific process or result. Another idea is raw data augmentation, which is done for image-based models by rotating or zooming in or out on the images, increasing the number of samples. This can also be applied to chemistry, for example, by writing multiple identical representations of the molecules. Neither case is as good as getting more data but sometimes that is all one can do.

For this case study, what were the extrapolation capabilities of these machine learning models? I also would like to hear your overall insight on the extrapolation of machine learning methods in general as well, i.e., when the maximum or minimum values are desired and are not available in the training set.

In the case study, the models worked well enough with look-ahead errors (i.e., on the unseen next round of complexes) around 0.2 eV, you can find all the relevant metrics and plots in the paper and SI. While not flawless, the predictions are of sufficient quality to help choose which complexes to study next, so I would say they are fit for purpose. Perhaps that is the best way to look at it as what is good enough for one application might not be sufficient in another. My general feeling is that the community is fairly good at predicting some properties (bandgaps, atomization energies) and worse at some others, such as biological activity (which is obviously a much more complicated quantity). In terms of not having representative training data, that is obviously not a good situation but I think it is almost more important to have some kind of uncertainty metric to warn that you are reaching too far away from training data since even compounds that lie in the range of our training points might have different chemistry and end up on the opposite end of the scale relative to where you would expect based on the training examples.

Is logP really a good surrogate for solubility? Did you consider alternative descriptors?

Possibly it is the best we can do. I think it is a reasonable option given the uncertainties for these systems. Unfortunately, our understanding of the solubility of inorganic complexes is not nearly as good as it is for organics and it is hard to find something unambiguously better that we can access with DFT. Since we depend on QM solvation energies, we at least account for buried vs. exposed polar groups in the ligands in the relevant complex assembly. However, I think that the main contribution of our work is a method that we believe is quite general, instead of a specific prediction or property.

DFT with approximate density functionals is well known to severely fail for high-spin open-shell cases, especially since, in these cases, the ground state electronic structures are multiconfigurational. Is this aspect usually ignored in your process of training the ANNs?

It is a good question and something that we worry about a lot. One way we probe the suitability of hybrid DFT is by routinely varying the fraction of exact exchange (Hartree-Fock exchange, HFX) used, from 0% (pure GGA) to 30%, and measuring how the relative spin state energies are perturbed. It turns out that different materials have different sensitivity to HFX, and we looked at predicting this sensitivity in previous work and even how we could use this information to steer our optimization away from complexes with predicted high sensitivity. Our ANN models can provide predictions at any HFX fraction desired. However, for transition metal complexes DFT will only ever take us so far, and my colleagues have been developing machine learning models to predict the extent of multireference character in complexes before simulating them, allowing us to deploy multireference methods automatically when we sample regions of chemical space that require them. Looking at their data, DFT is actually not all that bad on average though it is sometimes very wrong. This was not used in the case study I showed, so a sensible follow-up would be to screen any ideas with more accurate methods before actually making them. Still, it is a lot easier to do 30 multireference calculations instead of hundreds! For me, I am mostly interested in the optimization algorithm and I think it would be applicable to most things one could compute with DFT.

May I ask your opinion on the direct prediction of electron density?

This has been shown by Kieron Burke and Klaus-Robert Müller among others and there are also orbital-free DFT approaches that predict kinetic energy as a function of density. It looks very neat and could be highly transferable but it is difficult to obtain and manipulate reference data (3D point clouds are a pain). I think I prefer to model the actual endpoint directly, be that energy of the conformation since that is what we are actually trying to estimate since there is only one step in the process (input -> energy) and it is generally a little simpler.

Can you please say the exact properties you considered from DFT for ANN?

We only use DFT for the endpoints, i.e., redox potential and logP, so that we don’t need to do a DFT calculation for a new complex before making a prediction. We computed these quantities from free energy differences between complexes in different oxidation states and in different solvents

In your case study, you mentioned the product of nuclear charges (in shells) as a quantity to investigate. Is this something a human told the algorithms to try, or was this something the algorithm found that lead to better predictions of dG and/or log P?

A bit of both. We started with a large list of possibilities, inspired by our domain knowledge but also trying to be as inclusive as possible, and then used machine learning to select the most relevant variables. It actually depends on which output property you choose, for example, one gets a slightly different result when predicting spin state ordering vs predicting redox potentials. In that case, we could rationalize the finding from a chemical perspective, so we now do this analysis for any new properties to try and gain some insights into which part of the complex are the best targets for smart design (i.e., spin states are strongly controlled by the first shell, while the second shell still contributes a lot to redox potential). You can read all the details in the original paper here. The latest trend with graph convolutional neural networks is to essentially make this more automatic and require less domain input (since we might be wrong), but I think it depends on how much data you have, and especially for the smaller datasets human knowledge can add quite a bit.

Which programming languages are important for machine learning?

Python is the most common and has the most flexible, up-to-date libraries. Under the hood, the libraries call efficient functions written in C/Cpp (usually). But most languages have ‘good enough’ options for most tasks.

How many molecules did you use for training the model?

In the work I showed, I started with a few hundred (~300) up to a few thousand. For these systems, it is difficult to get much more, though organic chemistry datasets are usually a lot bigger.

Your training set was based on DFT, but how well does the DFT training set represent experimental data? How many hits from the machine learning predictions were observed in the lab, and how close were they to redox/solubility predictions?

Heather’s group only works on theory and the focus is to develop algorithms for combining quantum chemistry and machine learning for molecular optimization, so we don’t have the capacity to make the compounds in the lab. However, our approach has been compared to experiments and showed around a 0.3 eV difference. Numbers of solubility are much harder to come by! But I hope the approach is general could be applied to other systems as well.

How do you decide which descriptors to be included for machine learning model training?

The best practice is to try a number of different approaches and use a method such as cross-validation to select the descriptors that give you the lowest generalization risk. However, other factors can be important! If your descriptors are calculated by DFT or depend on experimentally measured properties (say melting point or lattice constant), you might run into a problem using your model on new cases where this data is not available. Also, we generally prefer descriptors that we can understand and explain, even if they give slightly worse performance than others that are harder to visualize and understand.

What functionals and basis set were you using?

We use mainly old-fashioned B3LYP + LANL2DZ, however, we run all of our calculations at a range of different Hartree-Fock exchange fractions from 0 (BLYP) to 0.3, since in our experience the fraction of exact exchange plays more of a role than the form of xc used. It is actually a really interesting thing to study as the sensitivity of spin state ordering to varying HFX – and we have built a model that actually predicts this sensitivity on a case-by-case basis. We have also worked extensively with predicting when DFT will not be appropriate and when we need to use a multireference method, so we can now make these modeling choices in an automatic way!


That is very cool work. In a sense enhanced sampling is related to the active learning I discussed, steering the model to a fruitful area in chemical/conformational space in a data-driven way. machine learning is being applied to every part of a chemical simulation, I look forward to running machine learning-metadynamics on a DFT-accurate neural network potential for a full protein in the near future!

Do you know about chemometrics, and what do you think about it?

It is not something I really work with or see used a lot, but I understand it as combining different measurements and signals to interpret chemical systems. I don’t really have an opinion about it but I have seen a lot of interest in getting machine learning methods more directly involved with understanding spectra (signals) directly, automated peak assignment, etc. I am not sure if you draw a distinction with cheminformatics, which to me is a little broader in scope involving QSAR, chemical database, etc. I think many of these methods use techniques that l would place under the umbrella of machine learning in chemistry in some sense, e.g. PCA or PLS, SVR. Certainly, people have been applying machine learning to chemistry for as long as there has been machine learning.

How can we deal with the lack of samples number if we can’t do more experiments or study more samples? Because machine learning needs a lot of samples to train and validate models. Is it possible to generate data or to use another approach?

There are a few neat ideas that can help. One idea I particularly like for chemistry is transfer learning, which is building a model to predict some kind of chemical property from a big database, and then fine-tuning it on some smaller dataset. For example, these authors trained a model on a large number of cheap(er)-to-compute DFT energies and then fine-tuned the model on more expensive coupled-cluster calculations. One can also do ‘unsupervised’ learning, where you train a model to predict some basic thing about the molecule such as hiding one atom or bond and training the model to supply the missing atom or bond label, as in this IBM work. In either case, you hope to teach the model the basic rules of chemistry and it is easier to teach it about some specific process or result. Another idea is augmentation, which is done for image-based models by rotating or zooming in or out on the images, increasing the number of samples. This can also be applied to chemistry, for example, by writing multiple identical representations of the molecules. Neither case is as good as getting more data but sometimes that is all one can do.

In, for example, PCA decomposition, can machine learning techniques provide any substantive advantages over traditional regression methods? I would say traditional regression methods are a subset of machine learning methods, and sometimes they are still the best choice, especially with limited data. That said, modern deep learning has definitely led to extreme advances in many areas, for example, image classification and natural language processing. In these fields, the difference in performance between traditional methods such as SVR and neural networks is large and indisputable. In chemical sciences, modern deep learning methods are allowing us to solve different kinds of problems compared to what we could do with traditional ‘shallow’ models, for example, neural network potentials, synthesis planning, and protein structure prediction. In these cases, they are not really competing with other methods as much as solving different problems. For some chemistry problems, such as QSAR, shallow methods are competitive or slightly worse than neural networks, though the difference is not large and we don’t really know why – it is a very difficult problem for machine learning because small changes (e.g. magic methyls) can lead to huge changes in the output, and usually the amount of data is not that much (the image classification people have millions of images to learn from). In regards to PCA specifically, as a dimensionality reduction technique, there are modern machine learning alternatives such as t-SNE or UMAP, which in my opinion tend to work better but there may be some cases where PCA is preferred, particularly because it can be easier to interpret and faster for large datasets. Since these methods are non-linear, they can recover patterns in the data that PCA cannot – here is a great online demo of t-SNE.

How do we design our training set?

A good training set should cover the intended chemical space where we will apply the model. In other words, it should be diverse and look like the data we want to predict. For example, if you are predicting the solubility of molecules, I would want to have a variety of polar groups and also some weakly soluble compounds. If your training data doesn’t include halogens, then you probably shouldn’t try and predict on data that contains them. Having good coverage of the space is more important than the raw number. Practically, one usually takes all the data one can get, but if you have the option to choose which data to collect, clustering and picking one from each cluster is probably a good idea.

What software is needed to do this type of machine learning?

I use pytorch, a python library for machine learning. Generally, there is a huge number of open source python packages for machine learning and many papers also make their code available. I have also historically used R, which is great for smaller datasets and simpler models but lags behind for complex neural networks.

Is there a principled or optimal way to identify the model family that will lead to the lowest true risk when properly fitted without having to train them all? Do you have advice for this?

In general, there is no way to know for sure which model will have the lowest (estimated) true risk without trying them all. However, you might be able to logically refine the search. For example, let us assume you fit 3 neural networks, one with 10 layers, one with 5 layers, and one with 2 layers. When you estimate the risk using cross-validation (CV), you determine that the best model has 2 layers, the next best has 5 and the 10 layer model is the worst since it is very overfitting (side note, this would manifest as the training error is near zero while the CV error is high). For your next attempt, it makes more sense to try models with 1, 3, or 4 layers instead of 20 or 200, since it seems like the 10-layer model is too complicated. In this way, you can iteratively refine the model search space, and this forms the basis of most hyperparameter search algorithms – a famous one is hyperopic. In practice, the search space is not 1D and the dimensions are NOT independent, but the idea is the same. They start with some random trials and then try and predict which areas of the search space are more productive. For complex models, this is the way to go, for simpler models and smaller search spaces a grid search is usually good enough.

Can you please suggest very basic books in machine learning for chemists?

An excellent machine learning book (not chem specific) that starts with the basics is Elements of Statistical Learning. That is where I started and I still use it! There is a pdf version on the Stanford website. Otherwise, the book Heather and I wrote is aimed at exactly this audience!

What Python libraries do you use the most?

PyTorch, RDKit, Pandas. I got a lot of use out of OpenBabel and Keras/Tensorflow previously.

How good do you need to be at math to learn machine learning?

It depends on what you want to learn. You can apply machine learning methods (and get good results) without really understanding how they work – many people who run computational chemistry calculations don’t necessarily know all the details of the simulations they run. However, getting a basic understanding of how machine learning works will obviously put you at an advantage. The good news is to understand and apply essentially all machine learning methods you only need some linear algebra and multivariable calculus. Even “how” neural networks work is mathematically simple – they are just collections of many neurons, each of which is pretty easy to understand. That said, there are more theoretical aspects such as how models generalize (or not), “why” large neural networks work so well, which involve some more complicated mathematics (mostly probability theory).

Can machine learning be used in all areas of chemistry, including inorganic and physical chemistry? What is the greatest difficulty in using it in a new area?

I think almost all scientific fields generate data in some form or another, and machine learning is really about if the patterns in that data can be used to predict future outcomes and so it can be applied in all these different applications. One issue in a new domain is lack of precedent as none of the existing descriptors or methods might represent your problem well – but this is also a great opportunity for domain knowledge to help set the machine learning problem up correctly! It can also be difficult to get enough data in specialized areas.

When the construction of metamodels that replace the usual numerical methods is implemented, what is the reduction in computation time in the calculations?

It depends on the model and what it is replacing. In the work I showed, I am swapping DFT calculations that take hours for machine learning models that take fractions of a second. To evaluate my full design space with DFT would have taken ~50 GPU-years, while the machine learning model takes ~4 seconds on a laptop. However, when comparing, for example, classical mechanics/force fields with neural networks, things are a lot closer. It is not my area but I believe the current neural network potentials tend to be slightly more expensive than force fields, although they give accuracy more comparable to DFT. An example paper where this is discussed is here. Some of the more complicated machine learning models could be considerably slower, but I don’t think there is any machine learning model that takes more than a few seconds per compound. Actually, classical machine learning methods such as SVR are often a lot slower than much larger neural networks, especially for large numbers of predictions, because neural networks typically scale linearly (while kernel inversion is cubic in principle).

Biomass decomposition is very complex. Can machine learning be used to carefully predict rate data of the decomposition of biomass, especially when we have about 50 data sets?

I doubt anyone knows – it depends on how the data points are represented and how ‘predictable’ they are. That sounds a bit vague so let me try to explain. One can usually fit Arrhenius rate expressions to reaction rates at a handful of temperatures. The reason is that the relationship between ‘x’ (inverse temperature) and ‘y’ (log rate) is very smooth (ideally linear). In the same way, the difficulty in predicting something comes down to how directly and smoothly the representation (x value) is related to the rate. For smaller data, how you represent them can be very important. Why not try it – especially with only 50 points you should probably use a very simple model, such as LASSO or ridge regression, and see what the results look like?

Compared with big data and other kinds of data, chemical data is a little bit small and expensive to enlarge. Which machine learning strategies can be employed to advance in this field?

While many types of chemical data are scarce, that is not always the case. There are publically available datasets of tens of millions of DFT calculations, millions of reactions mined from patents, and large databases like the CSD, materials project, open materials database, QOMD, PDB etc. Also, the amount of data needed can depend a lot on what you are trying to achieve – it may be that small data can be quite useful as long as it is similar to what you want to predict. However, in some cases, there is not enough data and one neat idea is transfer learning, i.e., building a model on one of these abundant data sources and then fine-tuning it on smaller datasets.

That was exactly the point – Jon uses terms such as minimal data and lots of data but it is hard to judge what minimal and lots actually mean. It is difficult to define how much data is ‘enough’ since different applications have different requirements and some datasets are just easier than others. As a toy example, consider fitting rate expressions to an Arrhenius plot (this is nothing other than a simple linear regression problem). You can probably get good kinetic parameters from a handful of reactions because the relationship between 1/T (the x variable) and log Rate (the y variable) is simple and smooth. It doesn’t need a very complex regression function to describe this data and so we don’t need much. In principle, the same applies to more complicated cases – if the way you represent your data, the x values, is smoothly correlated with what you are trying to predict, it is much easier. So it is not only a function of the amount of data but also how you represent it, and how accurate a model needs to be useful. In drug discovery, even a very weak model that is correct 10% of the time is very useful, since getting one active molecule for ten tries is a really good result. If you really want to put a number on things, I wouldn’t look too much into datasets with <50 points unless you can basically see the model by eye (as in the Arrhenius case), and I would only go to ‘deep’ models like neural networks with at least a few hundred (however, many classical algorithms such as LASSO could be very useful in that range).

Do you know of a community (Stack, Facebook group/page) to join and interact about machine learning?

I find the machine learning subreddit is very lively and even well-known people in the machine learning community such as Yoshua Bengio participate occasionally.

Is it possible to make predictions of free energy in reactions?

Yes, here is an example paper predicting barriers – they even make their trained model freely available so you could try it. In general, with enough data, we expect to predict any physical quantity. However, you can notice that they use quite a large quantum chemical dataset to train the model.

Sign-up for news and communications from ACS eBooks!

Meet the ACS ES&T Engineering Early Career Board

ACS ES&T Engineering is proud to announce the appointment of its first Early Career Advisory Board. The group collectively represents the full breadth of the best of the journal from excellent scholarship to leading cutting-edge research and beyond.

Take a few minutes getting to know some of our members of the ACS ES&T Engineering Early Career Board:

Marta Hatzell, Georgia Institute of Technology

Marta Hatzell

Marta Hatzell

What’s your background?

I have a background in mechanical engineering, environmental engineering, energy, and the environment.

What are you currently working on?

Broadly, we work on environmental-related catalysis and separations. More specifically, we investigate photocatalytic and electrocatalytic routes to capture and convert inert and waste nitrogen, electrified carbon capture and conversion, and systems central to the water-energy nexus.

What do you hope to bring to the journal?

Multi-disciplinary insights into the design and analysis of emerging environmental technologies.

What’s the most interesting challenge in your field at the moment?

Understanding and effectively taking social cost into consideration when designing and defining what is a sustainable technology.

Follow Marta’s research on Twitter: @MartaHatzell.


Amy V. Mueller, Northeastern University

Amy V. Mueller

Amy V. Mueller

What’s your background?

I have a background in electrical engineering, computer science, and environmental chemistry.

What are you currently working on?

Building better tools for studying the environment and optimizing environmental infrastructure, fusing science knowledge with designing new sensors, and leveraging data science innovation.

What do you hope to bring to the journal?

An interdisciplinary perspective around adapting “black box” numerical techniques for use in the sciences, especially around accuracy and trustworthiness issues.

What’s the most interesting challenge in your field at the moment?

The cost of collecting environmental data – meaning our datasets are likely always going to be “small!”


Ngai Yin Yip, Columbia University

Ngai Yin Yip

What’s your background?

My primary training is in physicochemical separation techniques for environmental applications, with a focus on water purification and desalination. I received my doctoral degree in chemical and environmental engineering from Yale University, where I worked on advancing novel membrane technologies for the sustainable production of energy and water. My postdoctoral research topic pivoted to microbiology and was performed at the Singapore Centre for Environmental Life Sciences Engineering. I received my undergraduate degree in civil and environmental engineering from Nanyang Technological University, Singapore.

What are you currently working on?

The overarching aim of my research is to develop and advance physicochemical technologies to address challenges at the interface of water, energy, and the environment. We are especially interested in energy-efficient desalination to address global water challenges. Our recent focus in that direction is on pioneering outside-the-box innovations to treat the emerging problem of hypersaline brines. We are also working on technological solutions for decentralized nutrient recovery from anthropogenic waste streams, which we think is a critical component for the transition to a circular economy. To achieve a principles-based approach for the development of next-generation membrane materials, we are investigating the fundamental phenomena governing transport in thin-film polymers. We hope the insights from our studies can guide the rational design and fabrication of membranes with enhanced separation selectivity.

What do you hope to bring to the journal?

I am excited to be a part of the ACS ES&T Engineering Early Career Board! I hope to bring an early-career perspective on the new directions and emerging challenges in water quality and quantity, contribute to elevating the scientific rigor of water engineering, and be a conduit between the journal and my fellow early career researchers.

What’s the most interesting challenge in your field at the moment?

While concentrating on developing engineering solutions for our water challenges, it is important not to lose sight of making the technologies and innovations as widely accessible as possible. With the inequality gap ever-widening, environmental engineers and scientists have an obligation to ensure that underprivileged communities, domestically and globally, do not end up being underserved because they are priced out of advances in water purification and treatment. We have much work to do to secure an equitable water future for all.

Learn more about Ngai’s research on his website.


Xin Wang, Nankai University

Xin Wang

Xin Wang

What’s your background?

I completed my B.E. in environmental engineering from Harbin Institute of Technology in 2004. After that, I continued my Ph.D. study under the supervision of Professor Yujie Feng in the same university, majoring in environmental science and engineering. During 2008-2009, I moved to Penn State University as a joint Ph.D. student, working with Professor Bruce Logan. My Ph.D. research focused on energy recovery from wastewater using microbial fuel cells. I joined the department of environmental engineering at Nankai University, China as a lecturer after I received my Ph.D. in the summer of 2010. I was promoted to associate professor and full professor in 2012 and 2017, appointed as the Department Chair of Environmental Engineering in 2019. I was a one-year visiting scholar at the University of Colorado Boulder, where I worked with Professor Zhiyong Jason Ren from 2015-2016. My research has been recognized by the Water-Star Prize (IWA-YWP China), Scopus Young Researcher Award, etc.

What are you currently working on?

My current research mainly focuses on the fundamentals of electroactive biofilm formation on the carbon electrode and the microbial ecology of electroactive and non-electroactive microbes. The electroactive microbes are considered as the energy transporter that can obtain electrons directly from the electrode to the accelerate biodegradation. I believe that the deep understanding of the relationship between these microbes will help us to develop novel biological augmentation technologies to solve environmental problems, such as the biodegradation of refractory pollutants and resource recovery from wastewater. Now we are developing microbial electrochemical technologies for fast treatment of organic pollutants in water and soil, active nitrogen recovery, dehalogenation and sensing biochemical oxygen demand (BOD), and early warning of toxicity.

What do you hope to bring to the journal?

I hope to bring my expertise on microbial electron transfer, biofilm formation and ecology, nutrient conversion and recovery from wastewater, and biosensing to ACS ES&T Engineering. I believe the interdisciplinary study of electrochemistry, microbiology, and ecology will be more and more important to push the conventional biodegradation technologies forward to a more efficient and controllable level in the future.

What’s the most interesting challenge in your field at the moment?

The understanding of microbial electron transfer in the biofilm, especially the interspecies extracellular electron transfer in complex biofilm formed from natural inocula (wastewater, soil, sediment), is interesting and challenging. In-situ investigation methods have to be developed to reveal the real-time response of electroactive microbes over time to the electrochemical parameters at both single-cell and microbial community levels. Molecular microbiological tools in combination with electrochemical and biological modeling are needed to fulfill the gap between theoretical and experimental studies. More effort should be devoted to linking the fundamental findings to new technology development. For the bioelectrochemical sensors, the design and standardization of electroactive biofilm, the data analysis with machine learning, and the device design are the challenges right now.


Heather Holmes, University of Utah

Heather Holmes

Heather Holmes

What’s your background?

My educational background is in mechanical engineering, but my research is transdisciplinary and incorporates environmental engineering, atmospheric science, and environmental health. I received my Ph.D. from the University of Utah with a focus on experimental atmospheric turbulence. After graduate school, I did two postdocs, one of which was at Georgia Tech, where I focused on air quality and exposure modeling for environmental epidemiology studies. In 2014, I started as an assistant professor in atmospheric science at the University of Nevada, Reno.

What are you currently working on?

In my research group in the Department of Chemical Engineering at the University of Utah, we are currently working in the areas of atmospheric turbulence, air pollution, and environmental health. Using ground-based sensors, atmospheric models, and satellite remote sensing, we investigate atmospheric physics, air pollution sources, transport and dispersion, and provide data for human health and public policy assessments. One aspect of our research is to improve atmospheric models for wildfire smoke transport and temperature inversions, where both frequently lead to poor air quality in the western U.S. We also use novel data analytics to incorporate multiple sources of data into estimates of human exposure to ambient air pollution.

What do you hope to bring to the journal?

I am excited to serve on the Early Career Board for ACS ES&T Engineering and to collaborate with a network of my peers. I also hope my knowledge and experience in air quality and environmental health brings strong interdisciplinary expertise to the growth of the journal.

What’s the most interesting challenge in your field at the moment?

There are several interesting challenges in the realm of atmospheric science and environmental health. The four challenges I spend the most time thinking about are:

  1. Better models for atmospheric turbulence to improve regional-scale numerical weather prediction modeling.
  2. Understanding personal-level human exposures to ambient and indoor air pollution.
  3. Fine-scale modeling of the spatiotemporal gradients of ambient air pollution concentrations that incorporate microscale atmospheric flows.
  4. How to provide big data learning opportunities and teach the future workforce to use open access datasets.

Xiaonan Wang, National University of Singapore

Xiaonan Wang

What’s your background?

I am an assistant professor in the Department of Chemical and Biomolecular Engineering at the National University of Singapore (NUS). I received my B.Eng. from Tsinghua University in 2011 and Ph.D. from University of California, Davis, in 2015. After working as a postdoctoral research associate at Imperial College London, I joined NUS as an assistant professor in 2017.

What are you currently working on?

My research focuses on the development of intelligent computational methods, including multi-scale modeling, optimization, control, data analytics, and machine learning for applications in energy, environmental, and manufacturing systems to support smart and sustainable development. I am leading a Smart Systems Engineering research group at NUS of more than 20 team members as PI and also the deputy director of the Accelerated Materials Development program in Singapore.

What do you hope to bring to the journal?

I hope to incorporate into the journal more state-of-the-art content on data-driven analysis and intelligent tools to benefit environment research and sustainable development. Low-carbon development to achieve carbon neutrality in the near future in multiple sectors is another direction I would like to develop. Moreover, I am encouraging all female researchers and students to be proud of our work and choose to challenge it.

What’s the most interesting challenge in your field at the moment?

How to obtain high-quality and large-amount big data for energy and environment systems.

Follow Xiaonan’s research on Twitter: @xnwang07.


Marina Eller Vance, University of Colorado Boulder

Marina Eller Vance

Marina Eller Vance

What’s your background?

I am an environmental engineer by training. I received my bachelor’s and masters at the Universidade Federal de Santa Catarina in southern Brazil and my Ph.D. at Virginia Tech. After obtaining my Ph.D., I had an unusual postdoctoral experience, where I helped run two research centers at Virginia Tech, one called the Virginia Tech Center for Sustainable Nanotechnology and one called NanoEarth, which is a node in the National Nanotechnology Coordinated Infrastructure. Despite all this focus on environmental nanotechnology, my primary field of expertise is actually air quality.

What are you currently working on?

My research is centered on applying engineering tools to better understand and minimize human exposure to environmental contaminants that may occur from everyday activities and the use of consumer products. My research group performs experimental investigations into the physical and chemical characteristics of aerosols from everyday sources, from emissions to subsequent transformations, in both indoor and outdoor environments.

What do you hope to bring to the journal?

I hope to contribute to this great new publication’s efforts in air quality and aerosol publications, both by performing peer reviews and aiding in other efforts that may enhance the contributions of diverse members of our scientific community.

What’s the most interesting challenge in your field at the moment?

Recently, my work has focused on indoor aerosols from everyday activities such as cooking. Indoor air quality, in general, is receiving renewed and much-deserved attention lately, in part due to the COVID-19 pandemic. The pandemic has brought two important issues front and center:

  1. The airborne transmission aspect of the coronavirus disease, bringing more attention to the need for ventilation and air cleaning, especially in schools and offices.
  2. The fact that people are spending more time at home, bringing more attention to air quality in the home. Both issues will have long-lasting effects on the way we do think about indoor air quality.

Follow Marina’s research on Twitter: @marinavance.


Kangwoo Cho, Pohang University of Science and Technology (Postech)

Kangwoo Cho

Kangwoo Cho

What’s your background?

I received a B.S. in economics and an M.S. in Civil, Urban, and Geosystem Engineering  from Seoul National University in South Korea. My research initiated from a biological nutrient removal process combined with electrochemical clarification of activated sludge. Thereafter, I joined the Korea Institute of Science and Technology (KIST) in 2006-2016 for the commercialization of environmental materials for filtration/adsorption-based water treatment. These experiences served as a momentum for the current works on electrochemically intensified environmental technologies. During 2010-2014, I pursued a Ph.D. at the California Institute of Technology under the supervision of Professor Michael R. Hoffmann. I was involved in the ‘Reinvent the Toilet’ project to work towards a self-standing water/energy cycle for sanitation facilities in developing countries, based on wastewater electrolysis cells. These solid backgrounds could establish my primary research field on environmental electrochemistry. My major honors include the environmental electrochemistry prize given by the International Society of Electrochemistry in 2015.

What are you currently working on?

My research interests nowadays span broadly in electrochemical or photochemical processes for water treatment coupled with energy conversion. In particular, the core principles in electrochemical water splitting, fuel cell, and battery could be modified for environmental purposes in terms of wastewater electrolysis/fuel cell and capacitive deionization, among others. The top priority is currently placed on material engineering to control electrochemical activity and the selectivity for generation of reactive oxygen/chlorine species, direct electron harvesting from pollutants, and non-Faradaic separation of ionic pollutants. These fundamental researches can be deployed for novel processes or intensification of existing practices.

What do you hope to bring to the journal?

Growing concerns on hitherto unforeseen aqueous micro-pollutants and renewable energy sources led us to variable electrochemical processes as the most direct methods to control and monitor redox transformation and separation of aqueous pollutants. However, the key requirements for the electrochemical water/energy nexus processes for broad applications would include proper engineering of the electrocatalysts with suitable properties in wastewater matrix. In addition, long-term stability should be secured for a practical application. The electrochemical water treatment processes have been commercially available for many years. However, results of a long-term field operation focusing on problems associated with side reactions in natural wastewater have been rarely shared in this community. As an early career board member of ACS ES&T Engineering, I want to contribute in terms:

  1. Suitability of electrocatalysts for wastewater treatment
  2. Sustainability of the electrochemical processes for a long-term operation, hopefully in a field condition.

What’s the most interesting challenge in your field at the moment?

Growing concerns on hitherto unforeseen aqueous micro-pollutants and renewable energy sources led us to variable electrochemical processes as the most direct methods to control and monitor redox transformation and separation of aqueous pollutants. However, the key requirements for the electrochemical water/energy nexus processes for broad applications would include proper engineering of the electrocatalysts with suitable properties in wastewater matrix. In addition, long-term stability should be secured for a practical application. Current issues in environmental and energy science should be addressed in terms of acceleration of elemental redox cycles. For example, oxidation of aqueous organic pollutants to carbon dioxide has been a long-standing aim of environmental science, whereas the reduction of carbon dioxide back into the energy carriers is paramount in energy science. In this regard, my research group seeks an engineering solution for the water-energy nexus based on the core principles of electrochemistry.

Learn more about Kangwoo’s research on his website.


Mingyang Xing, East China University of Science and Technology

What’s your background? 

I received my Ph.D. in Applied Chemistry from East China University of Science and Technology (ECUST) in 2012 and joined ECUST as an assistant professor in the same year. From 2015 to 2016, I worked as a visiting scholar in the research group of Professor Yadong Yin in University of California, Riverside. In 2019, I was hired as a professor by ECUST.

What are you currently working on? 

My current research interests include:

  1. The treatment of refractory organic pollutant wastewater by Fenton-like nanotechnology and other advanced oxidation technologies.
  2. Nano-photocatalysis combined with Fenton-like catalysis technology for the control of environmental pollutants.
  3. Reduction and recycling of CO2 molecules.

At present, as an independent PI, I lead a research group of about 20 team members and undertake research projects supported by the Natural Science Foundation of the Chinese Government. In addition, I also serve as the deputy director of Shanghai Engineering Research Center for Multi-media Environmental Catalysis and Resource Utilization.

What do you hope to bring to the journal? 

I am very happy to be an early career board member of ACS ES&T Engineering. I will be motivated to promote the application of nano- oxidation and reduction technology in the field of pollutant control and carbon neutralization.

What’s the most interesting challenge in your field at the moment?

  1. How to prepare nanosized catalysts at low cost on a large scale. 
  2. How to realize the combination of oxidation and reduction technology in the same reaction system. 
  3. How to find suitable application scenarios for nanotechnology.

Zhenfeng Bian, Shanghai Normal University

What’s your background?

I completed my Ph.D. in Environmental Chemistry from the Shanghai Normal University in 2010. After that, I was a JSPS Postdoctoral Fellow in the lab of Professor Tetsuro Majima during 2010–2013. In 2013, I started as a professor in the Department of Chemistry at Shanghai Normal University.

What are you currently working on?

My research interests are focused on the design and synthesis of nanomaterials for environmental photocatalysis. The main applications include precious metal recycling, heavy metal treatment, and pollutant degradation. I am leading a research group at SHNU of more than 20 team members as PI and am also the executive director of MOE Key Laboratory of Resource Chemistry.

What do you hope to bring to the journal?

I am excited to serve on the Early Career Board for ACS ES&T Engineering. Photocatalysis is a part of the frontier science of environmental chemistry. I hope to have a high-level journal in the field of photocatalysis, through which I can communicate with my peers around the world. It can help the technological research and industrial development of photocatalysis. It can also be used to popularize science and transform academic achievements.

What’s the most interesting challenge in your field at the moment?

There are still many scientific problems to be explored to realize the green and efficient recycling of valuable metals in solid waste by photocatalysis, so we still have a lot of work to do.


Shihong Lin, Vanderbilt University

What’s your background?

I am an environmental engineer by training and have research expertise in water separation and environmental interfacial science. I received my bachelor’s degree from Harbin Institute of Technology and my M.Sc. and Ph.D. degrees from Duke University, all in Environmental Engineering. I received my postdoc training at Yale University before joining Vanderbilt University in 2015.

What are you currently working on?

I am mostly interested in water separation processes for water treatment, desalination, and resource recovery. Our research group uses both experimental and theoretical approaches to study membrane, thermal, electrochemical, and hybrid separation processes to address challenges at the water-energy-food nexus. We are currently investigating the fundamentals and technologies for selective solute separation, energy efficient desalination, and brine management.   

What do you hope to bring to the journal?

I am very honored to be part of the inaugurating Early Career Board and hoping to contribute to the journal my expertise in water separation science and technology with the depth and breadth of my understanding on this topic. I also hope to serve as an ambassador for ACS ES&T Engineering and help it become the to-go-journal for high-quality papers in the field of environmental engineering and technology. 

What’s the most interesting challenge in your field at the moment?

Bridging the gap between fundamental scientific understanding and developing realistically impactful technological solutions.


Manish Kumar, Indian Institute of Technology Gandhinagar

 

What’s your background?

I have a background in environmental science and engineering with a specialisation in hydrogeobiochemistry.

What are you currently working on?

My research group’s overarching objective is to ascertain, broaden, comprehend, & develop various dimensions of the fate, transport, and remediation of geogenic, micro, microbial, and emerging contaminants in the freshwater systems. I am currently working on wastewater surveillance of COVID-19.

What do you hope to bring to the journal? 

A capability to deal with the article (that is submitted to ACS ES&T Engineering) of the interdisciplinary interface including hydrogeochemistry, water and soil pollution, metal speciation and toxicity, legacy/emerging contaminants and remediation technologies, and microbial and antibiotic resistance perspectives.

What’s the most interesting challenge in your field at the moment?

The top 10 trends are as follows:

  1. Application of wastewater surveillance for better pandemic preparedness through early warning system
  2. Treatment for SARS-CoV-2 RNA removal from the ambient environment: statistical and temporal significance
  3. Multi-drug resistance during and aftermath of the COVID-19 Pandemic
  4. Metal removal, partitioning and phase distributions in the wastewater and sludge: Performance evaluation
  5. Natural recharge and anthropogenic forcing imprints that influences Arsenic vulnerability
  6. Groundwater in-situ treatment perspective in the Post-COVID Anthropocene
  7. Utilization of Fly Ash Amended Sewage Sludge as Brick for Sustainable Building Material
  8. Perchlorate behavior in the context of black carbon and metal cogeneration
  9. Identification of aquifer-recharge zones and sources in an urban development area
  10. Understanding the extent of interactions between groundwater and surface water

Follow Manish’s research on Twitter: @manishkenv.


Drew Gentner, Yale University

What’s your background?

I am an Associate Professor of Chemical and Environmental Engineering at Yale University. Previously, I studied Environmental Engineering and Chemical Engineering at Northwestern (2007), and went on to the University of California, Berkeley for my M.S. and Ph.D. in Civil and Environmental Engineering (2012) where I had the valuable opportunity to work with fantastic researchers in environmental engineering, environmental sciences, and chemistry. I joined the Chemical and Environmental Engineering department at Yale in 2014, with a courtesy appointment in Yale’s School of the Environment.

What are you currently working on?

My research group focuses on air quality and atmospheric organic chemistry, with applications in both outdoor and indoor environments. This includes an emphasis on the emissions of reactive gas- and particle-phase organic compounds, especially from non-traditional sources (e.g. volatile chemical products) and their chemical transformations. We are passionate about analytical chemistry, with a focus on both accessible instrumentation to enable spatiotemporally-resolved studies as part of the SEARCH (Solutions for Energy, AiR, Climate, and Health) Center at Yale and high-resolution mass spectrometry methods with gas and liquid chromatography to elucidate the chemical composition of complex organic mixtures, their emissions, chemical-physical processes, and impacts.

What do you hope to bring to the journal? 

I am excited about ACS ES&T Engineering and thrilled to be part of the Early Career Board. My goal with my peers on the Early Career Board is to make the journal a venue for excellent research in environmental engineering and technology that addresses key issues in air pollution across outdoor and indoor environments and associated measurement technologies.

What’s the most interesting challenge in your field at the moment?

Emissions of pollutants and reactive chemical precursors, and the atmospheric chemical processes that all drive urban air pollution are evolving in fascinating ways across chemical, spatial, and temporal scales that present exciting challenges for measurement approaches, scientific understanding, modeling, and mitigation.

Follow Drew’s research on Twitter: @drew_gentner.

Call for Papers: Computational Advances in Protein Engineering and Enzyme Design

The Journal of Physical Chemistry B (JPC B) will publish a Virtual Special Issue (VSI) on “Computational Advances in Protein Engineering and Enzyme Design.” The VSI will be led by Guest Editors Professor Lynn Kamerlin of Uppsala University and Professor Etienne Derat of Sorbonne University. Together they encourage researchers to submit their new and unpublished work by Sept 15, 2021.

Research areas of particular interest include:

  • Enzymology
  • computational biochemistry
  • in silico protein design
  • bioinformatics
  • protein engineering

In conceiving this Virtual Special Issue, the Guest Editors were inspired by some recent exciting innovations and discoveries, including:

Enzymes are among the most complicated molecular objects and it was long thought it would be impossible to conceive a functional enzyme starting from scratch. Using natural enzymes to perform interesting chemical transformations has been feasible for a long time, but was historically limited to a range of specific reactions like alcohol oxidation. Extending the scope of unnatural enzymatic reactions was therefore a primary goal. Here, several ideas have been developed over the recent years: either using existing enzymes and extending their capabilities by exploiting their promiscuity, or adapting/adjusting a known active site to perform a new reaction. But, a more exciting challenge would be to develop ex nihilo an enzyme to catalyze a specific and new reaction. Here, computation is playing an increasingly important role, whether through identifying ancestral scaffolds that can be used as starting points for new enzyme engineering efforts, exploiting conformational dynamics for enzyme engineering using computational tools, exploiting machine learning approaches in enzyme design, or an explosion of online tools and web servers that can simplify the engineering process. The field is expanding and maturing rapidly, and taking on many new and different directions. The goal of this special issue will be to highlight computational advances in protein engineering and enzyme design, focusing on both methodology development and applications, and we welcome research articles or reviews from all broadly defined areas of this discipline.

Some articles already published as part of the VSI include:

Exploring the Minimum-Energy Pathways and Free-Energy Profiles of Enzymatic Reactions with QM/MM Calculations
J. Phys. Chem. B 2021, 125, 18, 4701–4713
DOI: 10.1021/acs.jpcb.1c01862

Introduction of a Glycine Linker Connecting the Heavy and Light Chains in Synthetic Cardosin B-Derived Rennet Changes the Specificity of Subpocket S3′
J. Phys. Chem. B 2021, 125, 17, 4368–4374
DOI: 10.1021/acs.jpcb.1c01826

Combined MD and QM/MM Investigations of Hydride Reduction of 5α-Dihydrotestosterone Catalyzed by Human 3α-Hydroxysteroid Dehydrogenase Type 3: Importance of Noncovalent Interactions
J. Phys. Chem. B 2021, 125, 19, 4998–5008
DOI: 10.1021/acs.jpcb.1c01751

Submission Instructions

The review process for all submissions for this VSI will be handled by JPC Senior Editor Benoit Roux.

To ensure an unbiased peer-review process, the journal asks that you do not indicate within your manuscript that the submission is intended for the VSI. If you do, your manuscript will be returned for correction. Instead, when you submit your manuscript, please indicate this on your cover letter and note what part and section you feel will be the best fit. You can find a complete list of sections and other important information for authors in the JPC Author Guidelines.

As with all submissions to JPC, your manuscript should represent a rigorous scientific report of original research, as it will be peer-reviewed as a regular article. Manuscripts are expected to provide new physical insight and/or present new theoretical or computational methods of broad interest.

Contribute to this Virtual Special Issue

If you are unsure if your research is within the VSI’s scope or have other questions about submitting a manuscript to this VSI, please email JPC B Deputy Editor Martin Zanni’s office at zanni-office@jpc.acs.org.

ACS Synthetic Biology Young Innovator Award Goes to Dr. Lei Stanley Qi

Dr. Stanley Qi is an assistant professor in the Department of Bioengineering, Department of Chemical and Systems Biology, and ChEM-H Institute at Stanford University. He obtained B.S. in Physics from Tsinghua University and Ph.D. in Bioengineering from the University of California, Berkeley. During his Ph.D., he studied synthetic biology with Dr. Adam Arkin and explored CRISPR biology with Dr. Jennifer Doudna.

After his Ph.D., he skipped postdoc training and performed independent research as a Systems Biology Faculty Fellow at UCSF. He joined Stanford in 2014. As a synthetic biologist, he developed the nuclease-dead dCas9 system, which greatly expands the CRISPR toolbox. Along with other researchers, he invented the CRISPR interference (CRISPRi) and CRISPR activation (CRISPRa) technologies for precise gene regulation. The work in his lab led to novel CRISPR tools for epigenome editing, 3D genome structure control, and DNA/RNA imaging. Using synthetic biology approaches, his lab developed CRISPR-based antivirals to treat SARS-CoV-2 and other RNA viruses. He won NIH Director’s Early Independence Award, Pew Biomedical Scholar, Alfred. P. Sloan Fellowship and an NSF CAREER Award.

I caught up with Dr. Qi to find out more about his career to date and what the ACS Synthetic Biology award means to him.

What does this award mean to you?

I am very excited to receive this award. When I was a graduate student thirteen years ago, I decided to change my major from physics to bioengineering after learning about the excitement in synthetic biology. It wasn’t an easy process, but I have enjoyed synthetic biology research. Although we still have a long way to solve many challenges, I am glad to see our efforts have received recognition from this award.

What are you working on now?

My lab has been developing CRISPR technologies as novel therapeutics to treat infectious diseases and regenerative medicine. It has been a dream for us to find safe and effective solutions to treat incurable diseases in medicine, and synthetic biology has become a key enabler. For example, using biomolecular engineering and gene design principles in synthetic biology, we developed CRISPR as a novel tool to precisely regulate many genes in human cells. We used these tools to switch on and off genes, which gives us new power to flexibly engineer the ‘inner DNA circuitry’ of cells, such as precisely converting human stem cells into neurons or creating better tumor-seeking and killing behaviors of immune cells.

How would you describe your research to someone outside your field of research?

CRISPR technology has become well known as a technological breakthrough to edit genes. We apply synthetic biology approaches to push the CRISPR technology to new limits beyond editing. For example, we help create a technology called CRISPR interference and CRISPR activation, known as CRISPRi and CRISPRa. CRISPRi/a interacts with DNA to silence or activate specific genes without altering the DNA sequence, which creates the ability to change stem cells into therapeutic cells such as neurons. We modify CRISPR as an imaging tool to capture real-time “movies” of the dynamic process of gene transcription and the movement of chromosomes. We also developed a CRISPR tool to engineer the 3-dimensional ‘DNA origami’ structure of the genome, to study human diseases related to the genome structure. Recently, we transformed the traditional CRISPR into an antiviral therapy to seek and destroy RNA viruses, which shows promise to treat COVID-19 and the flu.

What do you think is the biggest challenge currently in your area of research?

Biological research, in general, has remained a trial-and-error process. Synthetic biology, while trying to make biological research more predictable, designable, and ultimately programmable, still faces a lot of challenges. Take molecular engineering as an example. While synthetic biologists frequently need to engineer molecules (e.g., proteins derived from Nature or de novo designed) with new desired functions, this process has been tedious and sometimes unpleasant. We need new design principles, experimental approaches, and computational tools to make the biological design better. Fortunately, with the availability of powerful high-throughput synthetic biology approaches to generate huge amounts of data, either working or non-working data, as well as computational methods to analyze the data (e.g., machine learning), we see a future that can make ‘engineering biology’ easier and more predictable.

Have there been any highlights in your career to date that you are especially proud of?

I was lucky to witness the birth and development of the CRISPR field and was glad that I was able to make some contributions to the field. One research topic that I am proud of is the development of the nuclease dead Cas9 (which I named dCas9) in 2012, which was published in early 2013 when the CRISPR field was still in its infancy. The dCas9 system later became a basis and a platform for many RNA-guided applications and aided the development of CRISPRi/a, epigenome editing, chromatin imaging, base editing, and prime editing. It also facilitated the use of CRISPR as ‘wires’ to construct sophisticated circuits that were used to control metabolic flux, record cellular stimuli, and perform genetic screens. The development of the nuclease-dead dCas9 greatly expanded the toolbox of CRISPR for applications of genome engineering beyond nuclease-mediated gene editing.

Have you received any advice that you’ve found particularly helpful?

When I was a graduate student, I was curious how a fresh student like me can become an independent researcher. With this question, I asked my Ph.D. advisor, Dr. Adam Arkin. I clearly remember, he answered, “you should have a 30-year dream”. It wasn’t clear to me how this could help me become a better researcher at that moment. Later as I progress in my career, either by enjoying the exciting moments of discovery or ensuring the struggles from experimental failures, I start to realize how important it is to have a very long vision. Probably only with this type of vision, we will be less likely to be blinded by near-term successes or failures, nor get bored to progress to the next stop.

What would your advice be to someone just starting out in the field?

Keep your dream, but start with something simple. Keep exploring. There are so many exciting topics in synthetic biology, and I am almost certainly sure that the best of synthetic biology is yet to come.

View articles published by Dr. Qi in ACS Publications journals.

Dr. Qi will be presenting during the 2021 Synthetic Biology: Engineering, Evolution & Design (SEED) Meeting on Friday, June 18th. Join Dr. Qi virtually at 11:25 A.M. PDT for his talk ‘ACS Synthetic Biology Young Innovator Award: Synthetic biology for mammalian cell engineering and antivirals.’ See the full technical program here.

2021 ACS Macro Letters/Biomacromolecules/Macromolecules Young Investigator Award Winners

The ACS journals ACS Macro Letters, Biomacromolecules, and Macromolecules in partnership with the Division of Polymer Chemistry are proud to announce the selection of Professors Bradley D. Olsen of Massachusetts Institute of Technology and Haritz Sardon of POLYMAT/University of the Basque Country, Spain as the winners of the 2021 ACS Macro Letters/Biomacromolecules/Macromolecules Young Investigator Award. Professors Olsen and Sardon will be honored during an award symposium at the ACS Fall National Meeting, August 22 – 26, 2021.

2021 ACS Macro Letters/Biomacromolecules/MacromoleculesYoung Investigator Award Winners

Bradley D. Olsen, Massachusetts Institute of Technology

Professor Olsen was selected in recognition of his pioneering studies of protein-based polymeric materials and protein-polymer hybrids including the discovery of nucleoporin-like proteins (NLPs), block copolymers, and coacervates for applications in catalysis and biosensing; for the incorporation of topological defects in polymer network theory; and advancing line notation for data-driven polymer science through the invention BigSMILES built for polymer structures.

View selected articles published by Professor Olsen

Nucleopore-Inspired Polymer Hydrogels for Selective Biomolecular Transport
Biomacromolecules 2018, 19, 10, 3905–3916
DOI: 10.1021/acs.biomac.8b00556
***
Revisiting the Elasticity Theory for Real Gaussian Phantom Networks
Macromolecules 2019, 52, 4, 1685–1694
DOI: 10.1021/acs.macromol.8b01676
***
The Nature of Protein Interactions Governing Globular Protein–Polymer Block Copolymer Self-Assembly
Biomacromolecules 2014, 15, 4, 1248–1258
DOI: 10.1021/bm401817p
***
Adding the Effect of Topological Defects to the Flory–Rehner and Bray–Merrill Swelling Theories
ACS Macro Lett. 2021, 10, 5, 531–537
DOI: 10.1021/acsmacrolett.0c00909

Haritz Sardon, POLYMAT/University of the Basque Country

Professor Sardon was selected for this honor in recognition of his contributions to the development of organocatalyzed polymerizations and chemical recycling and upcycling of polymers. His accomplishments include preparing new functional polymeric materials using sustainable polymerization processes, such as synthesis of new polymers using “green” polymerization processes such as monomers from polymer recycling, reagents from renewable sources or the use of less hazardous organocatalysts

View selected articles published by Professor Sardon

Synthesis of Functionalized Cyclic Carbonates through Commodity Polymer Upcycling
ACS Macro Lett. 2020, 9, 4, 443–447
DOI: 10.1021/acsmacrolett.0c00164
***
Thioxanthone-Based Photobase Generators for the Synthesis of Polyurethanes via the Photopolymerization of Polyols and Polyisocyanates
Macromolecules 2020, 53, 6, 2069–2076
DOI: 10.1021/acs.macromol.9b02648
***
Isomorphic Polyoxyalkylene Copolyethers Obtained by Copolymerization of Aliphatic Diols
Macromolecules 2019, 52, 9, 3506–3515
DOI: 10.1021/acs.macromol.9b00469
***
Broad-Spectrum Antimicrobial Polycarbonate Hydrogels with Fast Degradability
Biomacromolecules 2015, 16, 4, 1169–1178
DOI: 10.1021/bm501836z