MolSSI Community Highlights spotlight exceptional examples of research and education enabled by MolSSI software and educational resources.

Dr. Danny Perez (Scientist IV in the Theory Division at Los Alamos National Laboratory) and his group focus on the development, implementation, and applications of long-timescale methods for atomistic simulation of materials through their work with the Exascale Atomistic Capability for Accuracy, Length, and Time (EXAALT) project. In this post, Dr. Perez tells us how he is using the MolSSI Driver Interface Project (MDI)  to unlock access to many different DFT software packages to  orchestrate the execution of very large numbers of calculations. Dr. Perez’s work has applications in the study of materials in extreme conditions and the generation of transferable datasets to parameterize machine learning potentials for materials.

MolSSI:

Hi Dr. Perez, we appreciate you taking the time to talk to us today! Your research field seems intriguing. Could you provide a brief introduction to your research area and share what your team is currently focusing on?

Dr. Perez:

My research focuses on the development, implementation, and applications of long-timescale methods for atomistic simulation of materials. I am especially interested in materials in extreme conditions, such as materials for fission and fusion energy, or particle accelerators. Recently, I became interested in the generation of transferable datasets to parameterize machine learning potentials for materials. I am also interested in applications of petascale and exascale computing in materials science.

MolSSI:

Can you share how MolSSI software tools or resources have been utilized to accomplish specific research objectives in your projects?

Dr. Perez:

I have collaborated with the MolSSI team as part of my role as PI in the Exascale Atomistic Capability for Accuracy, Length, and Time (EXAALT) project, which is one of the application projects supported by the DOE’s Exascale Computing Project. One of our objectives is to significantly decrease the time required to obtain reliable interatomic potentials for materials by integrating all steps, generation of training configurations, characterization with density functional theory, training of the potential, and validation. Our objective is to design scalable workflows to integrate and automate all of these steps.

MolSSI:

For this particular project, which MolSSI software tools or resources did your team employ, and what aspects or features did you find most beneficial?

Dr. Perez:

We are designing our workflows using MolSSI Driver Interface (MDI). We are especially interested in the plugin mode of MDI which enables us to control large numbers of instances of DFT and/or MD codes at scale using MPI without having to resort to file or socket-based mechanisms.

MolSSI:

How did MolSSI resources influence your software development efforts, and what led you to choose MolSSI resources for this purpose?

Dr. Perez:

In our case, we wanted to be able to orchestrate the execution of very large numbers of DFT calculations as part of an integrated workflow tool that can run at scale. The solution we converged to was to support engine codes that can be called through an API that can receive existing MPI communicators. Since such APIs are still not common or standardized, we faced the prospect of having to write specific code to support multiple engines. We quickly learned that the MolSSI MDI team aimed at standardizing such APIs, which would greatly simplify our work. After engaging with the MDI team, they endeavored to implement a new plugin mode which perfectly met our requirements in terms of integration with MPI codes at scale. This has led to a tremendous simplification of our code, as we only have to support MDI to unlock a whole ecosystem, instead of having to write engine-specific code.

MolSSI:

It’s fantastic to see the role MolSSI’s software has played in your project’s development! Could you share any scientific findings or results that your workflows have facilitated?

Dr. Perez:

We are still developing our workflow, but we hope to demonstrate the end-to-end generation of reliable interatomic potentials in under a day.

MolSSI:

Those are impressive goals, and it’s wonderful to see how MolSSI resources have contributed to your group’s success. Thank you for sharing your experiences and insights with us!

Although there are no publications yet that incorporate MolSSI software tools or resources in Dr. Perez’s work, we look forward to seeing the advancements his group will make in the near future.

MolSSI Community Highlights spotlight exceptional examples of research and education enabled by MolSSI software and educational resources.

Dr. Ryan C. Fortenberry, Associate Professor of Chemistry at the University of Mississippi, and his group focus on Theoretical Astrochemistry. They use computers to simulate the way light interacts with molecules and how molecules interact with each other in space. They have utilized MolSSI software in their work with the MOPAC program, an open-source semiempirical thermochemistry software for calculating molecular properties. MOPAC has been developed as commercial software for the last 30 years, and MolSSI has facilitated its transition to an open-source software project that provides it with the opportunity to be maintained and further developed by a community of open-source software developers. In this post, Prof. Fortenberry shares with us how they are leveraging the MolSSI software to reparameterize semi-empirical methods for the computation of anharmonic vibrational frequencies.

MolSSI:

Can you give us a short summary of your research area and what your group is currently working on?

Professor Fortenberry:

My group focuses on Theoretical Astrochemistry. We use computers to simulate the way light interacts with molecules in space and how molecules interact with each other in space. Currently, we have been working to reparameterize semi-empirical methods for the computation of anharmonic vibrational frequencies for the prediction of IR spectra for polycyclic aromatic hydrocarbons (PAHs) with application to ongoing observations from the James Webb Space Telescope (JWST).

MolSSI:

Could you tell us how MolSSI software tools or resources have been employed to meet specific research objectives in your projects?

Professor Fortenberry:

The Basis Set Exchange is used in all of our projects, as the ready translation between quantum chemical packages for basis sets within the exchange is the easiest means of importing new basis sets.  

The MOPAC program is one of the most readily accessible quantum chemical codes that includes semi-empirical methods. Changes to the house-made programs that utilize MOPAC as part of the workflow are able to be run on local computers (rather than only on supercomputer clusters) and even integrated with the GitHub continuous integration features thanks to the fact that MOPAC is now open-source and easily installable.

MolSSI:

Can you tell us about what features of MolSSI’s software were the most useful to you?

Professor Fortenberry:

While semi-empirical computations are available in numerous quantum chemical codes, the open-source nature of MOPAC allows for numerous additions that are unique to our applications. My student, Brent Westbrook, has been able to contribute code to MOPAC to be able to utilize new semi-empirical parameters as part of the input file instead of requiring a separate external file. MOPAC’s flexibility also allows the most useful number of significant digits for the PAH applications to be printed in the output file.  This has allowed Brent to develop an exceptionally streamlined protocol for optimizing the parameters for our PAH IR frequency computations for JWST.

MolSSI:

It’s amazing to see the role MolSSI’s software has played in your project’s development! Could you share any scientific findings or results from your research?

Professor Fortenberry:

Quartic force fields (QFFs) are fourth-order Taylor series expansions of the potential portion of the Watson Hamiltonian.  While sparse compared to a global potential surface, they still scale geometrically with the number of atoms.  Hence, the use of these as a means of computing anharmonic vibrational frequencies has typically been limited to small molecules.  The use of MOPAC and reparameterized semi-empirical methods trained thereon has enabled us to compute fully anharmonic spectra of PAHs of up to 3 rings (~25 atoms; on a laptop!) in preliminary tests with 10 rings a distinct possibility. Our preliminary tests indicate that we may be able to handle PAHs of up to 10 rings.

MolSSI:

Those sound like significant advancements, and it’s fantastic to see how MolSSI resources have helped your research. Thank you for taking the time to share your experiences and insights with us!

If you’d like to read more about Professor Fortenberry’s work, we encourage you to attend the upcoming MOPAC user group meeting and to see the following links and publications.

(1)        Westbrook, B. R.; Layfield, J. P.; Lee, T. J.; Fortenberry, R. C. Reparameterized Semi-Empirical Methods for Computing Anharmonic Vibrational Frequencies of Multiply-Bonded Hydrocarbons. Electron. Struct. 2022, 4 (4), 045003. https://doi.org/10.1088/2516-1075/aca458.

(2)        Westbrook, B. R.; Fortenberry, R. C. Pbqff: Push-Button Quartic Force Fields. J. Chem. Theory Comput. 2023, 19 (9), 2606–2615. https://doi.org/10.1021/acs.jctc.3c00129. Cover Article. GitHub repository: https://github.com/ntBre/pbqff

MolSSI Community Highlights spotlight exceptional examples of research and education enabled by MolSSI software and educational resources.

Prof. Jay Foley (Associate Professor of Chemistry at University of North Carolina, Charlotte) and his group have harnessed the power of MolSSI Educational resources to transform their approach to scientific software development. In this post, Prof Foley tells us how MolSSI tools and workshops have played a role in advancing his group’s research on light-matter interactions and computational design of nanomaterials for energy applications.

MolSSI: 

Hi Professor Foley, thanks for joining us today! Your research area sounds fascinating. Can you give us a quick overview of your research area and what your group is currently working on?

Professor Foley: 

My group is interested in the theory and modeling of light-matter interactions, and the computational design of materials for energy.  We are developing new computational methods that combine tools of ab initio electronic structure theory with cavity quantum electrodynamics to provide tools to simulate molecules strongly interacting with light (molecular polaritons), and we are also interested in developing and using computational tools based on classical electrodynamics to design nanomaterials with tailored optical and thermal radiative properties for energy applications.

MolSSI:

That sounds both interesting and impactful. Can you describe how you have utilized MolSSI software tools or resources to achieve specific research goals?

Professor Foley:

We have developed an open-source software package called WPTherml (Wicked Python package for Thermal Energy and Radiation management with Multilayer nanostructures) that couples rigorous electrodynamics computations to thermal radiation equations and aims to provide a powerful computational design engine for multilayer nanostructures for applications where control of optical and/or thermal radiation properties are paramount.  Some applications of particular interest to my group include passive radiative cooling, solar thermophotovoltaics, efficient incandescent lighting, and polaritonics.

We used the Software Development Best Practices workshop and the Objected Oriented Programming and Design Patterns workshop.  We participated in a virtual Software Development Best Practices workshop led by Dr. Jessica Nash in Summer 2021,  and we worked independently through some of the OOP and Design Patterns workshop material.  We have also used the MolSSI CookieCutter project. 

MolSSI:

Can you tell us a little more about the development of WPTherml? Why did you decide to use MolSSI resources, and how did they enable your software development efforts?

Professor Foley:

The WPTherml project was started around 2018, and at the time my group did not really have any experience with software package development.  Our initial goal for the package was fairly modest – we were utilizing electrodynamics computations (specifically, using transfer matrix method) to model thermal radiation in multi-layer planar materials, and we wanted to make these computations easy to use for students regardless of their experience level with computation and regardless of their computing platform. 

Over time, we wanted to add additional capabilities to this package – we wanted to add more electrodynamics solvers for different types of nanomaterial structures, add some quantum mechanical models to enable multi-scale modeling, and also add optimization capabilities to enable some elements of materials design. The structure of the original WPTherml package did not make these developments particularly easy, and we decided we should probably rethink and rewrite it to really make progress.  

We contacted MolSSI (specifically Jessica Nash) about participating in the Software Development Best Practices workshop, and she organized a virtual workshop for us.  We went through setting up a prototypical package (MolSSI’s CookieCutter) in that workshop with a lot of helpful structure built-in: documentation with sphinx, automated testing using pytest and GitHub actions, and also learned how to improve our git workflow.  We utilized the CookieCutter structure as the base for the new version of WPTherml.  

After the conclusion of that workshop, we spent a few weeks going through MolSSI’s Object Oriented Programming and Design Patterns workshop material on our own and decided to utilize the factory design pattern for WPTherml, since it seemed to naturally fit with our desire to utilize different types of physics solvers (e.g. transfer matrix method, generalized Mie theory, several quantum mechanical models) to produce similar types of outputs (spectra) from different structure classes.  

MolSSI:

It’s great to hear that MolSSI’s resources played an important role in the development of your software! Can you now tell us about what scientific findings your software has enabled?

Professor Foley

We have used the WPTherml package to design spectrally selective thermal emitters with record-breaking performance for solar thermophotovoltaics that were then fabricated and characterized by experimental collaborators, and have also used it to elucidate the optical properties of borophene (monolayer of boron).  

MolSSI:

Those are impressive achievements, and it’s wonderful to see how MolSSI resources have contributed to your group’s success. Thank you for sharing your experiences and insights with us!

If you’d like to read more about Professor Foley’s work, we encourage you to see the following links and publications.

WPTherml’s GitHub page: https://github.com/FoleyLab/wptherml

(1)        Suchanek, F.; Varner, J.; Lakatos, A.; Bello, J.; Soufanati, S.; Foley, J. J. I. Embracing Modern Software Development Best Practices in an Undergraduate Research Setting: A Case Study with the WPTherml Software Package. In Physical Chemistry Research at Undergraduate Institutions: Innovative and Impactful Approaches, Volume 1; ACS Symposium Series; American Chemical Society, 2022; Vol. 1428, pp 39–52. https://doi.org/10.1021/bk-2022-1428.ch003.

(2)        Varner, J. F.; Wert, D.; Matari, A.; Nofal, R.; Foley, J. J. Accelerating the Discovery of Multilayer Nanostructures with Analytic Differentiation of the Transfer Matrix Equations. Phys. Rev. Res. 2020, 2 (1), 013018. https://doi.org/10.1103/PhysRevResearch.2.013018.

(3)        Jeon, N.; Mandia, D. J.; Gray, S. K.; Foley, J. J. I.; Martinson, A. B. F. High-Temperature Selective Emitter Design and Materials: Titanium Aluminum Nitride Alloys for Thermophotovoltaics. ACS Appl. Mater. Interfaces 2019,11 (44), 41347–41355. https://doi.org/10.1021/acsami.9b13944.

(4)        Varner, J. F.; Eldabagh, N.; Volta, D.; Eldabagh, R.; Iv, J. J. F. WPTherml: A Python Package for the Design of Materials for Harnessing Heat. 2019, 7 (1), 28. https://doi.org/10.5334/jors.271.

(5)        Jeon, N.; Hernandez, J. J.; Rosenmann, D.; Gray, S. K.; Martinson, A. B. F.; Foley IV, J. J. Pareto Optimal Spectrally Selective Emitters for Thermophotovoltaics via Weak Absorber Critical Coupling. Advanced Energy Materials 2018, 8 (25), 1801035. https://doi.org/10.1002/aenm.201801035.

Join Molecular Sciences Software Institute (MolSSI) and Intel for the upcoming hands-on Basics of Accelerated Computing with Intel OpenMP GPU Offload virtual workshop:

• Registration closed.
• Level: Intermediate
• Date: Friday May 12 at 11:30-15:30 ET
• Recording

You can stay up-to-date about the upcoming workshops by subscribing to our newsletter here or following the updates posted on our Industrial Training Program webpage.

The Molecular Sciences Software Institute (MolSSI) is pleased to announce that the 2023-24 Software Fellowship competition will be open for submissions for ONE-YEAR Fellowships (from 1 July 2023 thru 30 June 2024) on February 15. These prestigious Fellowships recognize advanced graduate students and postdocs pursuing the development of software infrastructure, middleware, and frameworks that will benefit the broader field of computational molecular sciences, including biomolecular and macromolecular simulation, quantum chemistry, and materials science. Recipients will have the opportunity to work with scientists at the MolSSI in order to implement recommended best practices, putting the Fellow’s project on a firm foundation. Fellows will receive specialized training in state-of-the-art software design principles and tools, and they will engage in outreach and educational efforts organized by the MolSSI.  Each Fellow will be assigned a mentor among the Institute’s Software Scientists, who will oversee their software development efforts and training.  In addition, the 2023-24 Fellows will spend a week at MolSSI HQ (typically in late July or August; Sunday PM through Friday PM) to interact with the Software Scientists during their “Software Best Practices Boot Camp.”

Inquiries to: fellowships@molssi.org

The MolSSI Software Fellowships will include a number of benefits:

Graduate Student Fellowships:

• A generous stipend of $40k total for the 12-month Fellowships, plus tuition and required fees;
• Travel allowances sufficient to fully cover visits to the MolSSI headquarters, as well as participation in MolSSI workshops or summer schools;
• Access to MolSSI computational resources.

Postdoctoral Fellowships:

• A generous stipend of $50k total for the 12-month Fellowship;
• Travel allowances sufficient to fully cover visits to the MolSSI headquarters, as well as participation in MolSSI workshops or summer schools;
• Access to MolSSI computational resources.

Note that MolSSI Fellowships are paid from NSF Participant Support funds. The only allowable expense categories are stipend, tuition and fees; F&A (indirect costs) and health insurance costs are not allowed.

Eligibility and Selection Criteria:

The MolSSI Software Fellowship program is presently limited to graduate students and postdoctoral associates at accredited U.S. academic institutions. Graduate student Fellows are expected to be enrolled full time at their home university, and postdocs must be fully employed or have an offer of full employment prior to acceptance of the award.  The MolSSI Software Fellowships are an equal opportunity program open to all qualified persons without regard to race, gender, religion, age, physical disability or national origin.  Awardees may accept other concurrent awards that do not require a significant time or service commitment.

MolSSI Software Fellows will be selected by the MolSSI Science and Software Advisory Board based on (1) the quality of the applicant’s software research proposal and its relevance to the Institute’s goals; (2) the applicant’s research productivity, including previous software-development efforts; (3) previous academic performance; and (4) external references.

Application Requirements:

A complete MolSSI Software Fellowship Application must include:

• Project description (maximum of two pages, not including references);
• Current curriculum vitae;
• Undergraduate transcripts (graduate applicants only); graduate transcripts (all applicants);
• Examples of previous software development products (e.g. links to source-code repositories or other code samples);
• Two letters of support, one of which must come from the applicant’s current research adviser;

Please complete the application and upload the required documents HERE.
Letters of recommendation should be uploaded separately by your references HERE.

To receive full consideration for a Software Fellowship, all application materials (including letters of recommendation) must be submitted by April 1, 2023.

Additional Information:

• Award Number (FAIN): 2136142
• Award Instrument: Cooperative Agreement
• Award Date: 07/30/2021
• Award Period of Performance: Start Date: 08/01/2021 End Date:
  07/31/2026
• Project Title: S2I2: Impl: The Molecular Sciences Software Institute
  Managing Division Abbreviation: CHE
• Research and Development Award: Yes
• Funding Opportunity: NSF 20-1 Proposal & Award Policies & Procedures
  Guide – PAPPG
• CFDA Number and Name: 47.049 Mathematical and Physical Sciences

The workshop schedule is now available at THIS LINK!

Organizers:
Pratyush Tiwary, University of Maryland, College Park
Cecilia Clementi, Free University Berlin
Teresa Head-Gordon, University of California, Berkeley

In recent years, the field of machine learning (ML) has seen an incredible and sustained surge in interest. From image classifiers to board games to protein structure prediction, ML and big data and internet are making large impacts in nearly every field, including chemistry and molecular sciences. In 2019 MolSSI organized a ML and chemistry workshop at which point perhaps one could have been skeptical about the impact of ML in chemistry. With the passage of three years, the likely number of ML skeptics in chemistry is much smaller. A lot more has happened since 2019 and there is an urgent need to assess where we are as a community, as well as establish the software needs and directives.

This workshop, which will be held from May 31 to June 2, 2023, at the University of Maryland, will look at the following questions/areas through invited talks from industry/academia experts, brainstorming sessions and poster sessions:

(a) How to implement public access to reliable benchmark data to enforce reproducibility.
(b) What software tools & infrastructures are missing for chemists to use ML in everyday work?
(c) How to train the next generation of chemistry students who can use such ML software.
(d) Open-source standards that need to be set and met, which can often be challenged by the intellectual property rights related to these novel and important approaches.
(e) Discussions on problems faced in general in Machine Learning for Chemistry.
(f) Machine Learning for Chemistry from an industry perspective

A limited number of students/postdoc/additional faculty will be accepted for participation in the workshop. Preference will be given to those presenting posters; if accepted, they will be provided partial reimbursement for travel and accommodation.

Should you have additional questions about content or other aspects of the workshop, contact Prof. Pratyush Tiwary at ptiwary@umd.edu.

List of invited speakers:

1. Robert Abel, Schrodinger
2. Luigi Bonati, IIT Genoa
3. Rose Cersonsky, UWMadison
4. Bingqing Cheng, IST Austria
5. Connor Coley, MIT
6. Roberto Covino, Frankfurt Institute for Advanced Studies
7. Elizabeth Decolvenaere, D. E. Shaw Research
8. Amir Barati Farimani, Carnegie Mellon University
9. Andrew Ferguson, University of Chicago
10. Rafael Gomez-Bombarelli, MIT
11. Mojtaba Haghighatlari, Pfizer
12. Shantenu Jha, Rutgers University
13. Tyler Josephson, University of Maryland Baltimore County
14. Rajat Maji, UCLA
15. Simon Olsson, Chalmers University of Technology
16. Francesco Paesani, UCSD
17. Guido Falk von Rudorff, University of Kassel
18. Grant Rotskoff, Stanford University
19. Patrick Sahrmann, University of Chicago
20. Sapna Sarupria, UMinnesota
21. Mark Tuckerman, NYU
22. Omar Valsson, University of North Texas
23. Andrew White, University of Rochester