GSoC Ideas 2023: Difference between revisions
| (9 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
==Guidelines== | ==Guidelines== | ||
Open Chemistry is an umbrella for projects in chemistry, materials science, biochemistry, and related areas | Open Chemistry is an umbrella for projects in chemistry, materials science, biochemistry, and related areas. | ||
One important factor is that GSoC in 2023 will include both shorter projects (~175 hours) and longer projects (~350 hours). You should consider the appropriate timeline for your project proposal. We have indicated in the project totals where we suggest particular lengths. | One important factor is that GSoC in 2023 will include both shorter projects (~175 hours) and longer projects (~350 hours). You should consider the appropriate timeline for your project proposal. We have indicated in the project totals where we suggest particular lengths. | ||
| Line 7: | Line 7: | ||
Contributors can also decide on the number of weeks (e.g., spreading the project time over multiple weeks). | Contributors can also decide on the number of weeks (e.g., spreading the project time over multiple weeks). | ||
If you are unsure of the scope of a project, please reach out and discuss BEFORE the proposal deadline. | If you are unsure of the scope of a project, please reach out and discuss BEFORE the proposal deadline (20 March to 4 April). | ||
When possible, submitting drafts a week or more in dance of the proposal deadline is preferred because we can make suggestions towards your proposal. | When possible, submitting drafts a week or more in dance of the proposal deadline is preferred because we can make suggestions towards your proposal. | ||
| Line 44: | Line 44: | ||
'''Mentor:''' Geoff Hutchison (geoffh at pitt dot edu) or Marcus D. Hanwell (mhanwell at bnl dot gov) | '''Mentor:''' Geoff Hutchison (geoffh at pitt dot edu) or Marcus D. Hanwell (mhanwell at bnl dot gov) | ||
=== Project [175 or 350 hours]: Tools for Interactive Molecular Dynamics === | === Project [175 or 350 hours]: Tools for Interactive Molecular Dynamics === | ||
| Line 75: | Line 65: | ||
'''Mentor:''' Geoff Hutchison (geoffh at pitt dot edu) | '''Mentor:''' Geoff Hutchison (geoffh at pitt dot edu) | ||
=== Project [175 or 350 hours]: Improved Selection and/or Molecular Find-and-Replace === | |||
'''Brief explanation:''' Improved support for selection tools (e.g., "select all water molecules" and "select everything within X Å of my mouse-click") and tools for bulk editing of structures using a "find and replace" interface (e.g., replace 10% of all gold atoms with silver or "change this molecular pattern to this new functional group") | |||
'''Expected results:''' A set of new selection tools and commands and/or a set of tools to substitute atoms and molecular fragments in the builder. For example, the Wilmer group published "MOFUN" (https://doi.org/10.1039/D2DD00044J) a package to replace specific fragments, and many packages such as Open Babel and RDKit support SMIRKS or reaction SMILES rules. Ideas to improve the usability are highly welcome. | |||
'''Prerequisites:'' Experience in C++, some experience with Python will be helpful. | |||
'''Mentor:''' Geoff Hutchison (geoffh at pitt dot edu) | |||
=== Project [175 hours]: Integrate with RDKit === | |||
'''Brief explanation:''' Integrate the RDKit toolkit into Avogadro for conformer sampling and force field optimization | |||
'''Expected results:''' RDKit is a BSD-licensed cheminformatics toolkit with a wide range of features useful for Avogadro 2. Most notably, RDKit offers efficient and accurate 3D coordinate generation, conformer sampling, and force field optimization. Implement a connection between Avogadro objects (molecules and atoms) and RDKit objects and implement conformer sampling and force field optimization code. | |||
'''Prerequisites:'' Experience in C++, some experience with Python will be helpful. | |||
'''Mentor:''' Geoff Hutchison (geoffh at pitt dot edu) | |||
==Open Babel Project Ideas== | ==Open Babel Project Ideas== | ||
| Line 123: | Line 133: | ||
Such a model could be used as a filter, or as a warning to flag up problematic structures. | Such a model could be used as a filter, or as a warning to flag up problematic structures. | ||
To be clear "model" does not necessarily mean a machine learning model, instead as a set of filters. (An ML-based model for tautomers might be useful, however.) | |||
Code could be modeled on MolVS using RDKit [[https://molvs.readthedocs.io/en/latest/]] | Code could be modeled on MolVS using RDKit [[https://molvs.readthedocs.io/en/latest/]] | ||
| Line 299: | Line 311: | ||
=== Project: [350 hours] Extended Interoperability of Denspart to Open Force Field Software === | |||
'''Brief Explanation: '''Partitioning a chemical system and its various properties into atomic contributions is not only qualitatively useful for chemical analysis but important for quantitative computational chemistry (e.g., charges in molecular mechanics force fields). This can be done by partitioning the molecular electron density into fuzzy atomic densities using Denspart ( package and class within ChemTools package), but there are alternative approaches based on topological partitioning of molecular density or partitioning of molecular orbitals. The goal of this project is to develop a unified framework determining the molecular contributions of specified atoms and/or functional groups. | '''Brief Explanation: '''Partitioning a chemical system and its various properties into atomic contributions is not only qualitatively useful for chemical analysis but important for quantitative computational chemistry (e.g., charges in molecular mechanics force fields). This can be done by partitioning the molecular electron density into fuzzy atomic densities using Denspart ( package and class within ChemTools package), but there are alternative approaches based on topological partitioning of molecular density or partitioning of molecular orbitals. The goal of this project is to develop a unified framework determining the molecular contributions of specified atoms and/or functional groups. | ||
| Line 315: | Line 327: | ||
'''Mentors:''' Farnaz Heidar-Zadeh (farnaz.heidarzadeh at queensu dot ca), Toon Verstraelen (Toon.Verstraelen at ugent dot be), and Esteban Vohringer-Martinez (estebanvohringer at qcmmlab dot com). | '''Mentors:''' Farnaz Heidar-Zadeh (farnaz.heidarzadeh at queensu dot ca), Toon Verstraelen (Toon.Verstraelen at ugent dot be), and Esteban Vohringer-Martinez (estebanvohringer at qcmmlab dot com). | ||
=== Project: [350 hours] Extended Interoperability of IOData to Open Force Field Software === | |||
'''Brief Explanation: '''IOData () is a free and open-source Python library for parsing, storing, and converting various file formats commonly used by quantum chemistry, molecular dynamics, and plane-wave density-functional-theory software programs. It also supports a flexible framework for generating input files for various software packages. The goal of this project is to extend the IOData framework to describe a chemical system for molecule mechanics (MM) modeling, specifically in the context of non-covalent interactions of open force-fields. | '''Brief Explanation: '''IOData () is a free and open-source Python library for parsing, storing, and converting various file formats commonly used by quantum chemistry, molecular dynamics, and plane-wave density-functional-theory software programs. It also supports a flexible framework for generating input files for various software packages. The goal of this project is to extend the IOData framework to describe a chemical system for molecule mechanics (MM) modeling, specifically in the context of non-covalent interactions of open force-fields. | ||
| Line 331: | Line 343: | ||
'''Mentors:''' Farnaz Heidar-Zadeh (farnaz.heidarzadeh at queensu dot ca), Toon Verstraelen (Toon.Verstraelen at ugent dot be), Esteban Vohringer-Martinez (estebanvohringer at qcmmlab dot com). | '''Mentors:''' Farnaz Heidar-Zadeh (farnaz.heidarzadeh at queensu dot ca), Toon Verstraelen (Toon.Verstraelen at ugent dot be), Esteban Vohringer-Martinez (estebanvohringer at qcmmlab dot com). | ||
=== Project: [350 hours] Deriving Atomic and Bond Properties from Molecule Orbital Partitioning === | |||
'''Brief Explanation: '''In traditional orbital-based atomic partitioning methods for molecular wavefunctions, one observes that the results are strongly dependent on the molecular basis set, and tend to work poorly for large atom-centered basis sets and be inapplicable to delocalized (e.g., plane wave) basis functions. One way to avoid these issues is using various sorts of quasi-atomic orbitals (QUAOs), including quasi-atomic molecular basis orbitals (QUAMBO), intrinsic atomic orbitals (IAOs), and other choices. Another alternative is to use (approximate) orbital-based equivalencies to spatial decomposition methods like the Hirshfeld or Bader partitioning. The goal of this project is to implement these approaches in the OrbsTools module of ChemTools (). | '''Brief Explanation: '''In traditional orbital-based atomic partitioning methods for molecular wavefunctions, one observes that the results are strongly dependent on the molecular basis set, and tend to work poorly for large atom-centered basis sets and be inapplicable to delocalized (e.g., plane wave) basis functions. One way to avoid these issues is using various sorts of quasi-atomic orbitals (QUAOs), including quasi-atomic molecular basis orbitals (QUAMBO), intrinsic atomic orbitals (IAOs), and other choices. Another alternative is to use (approximate) orbital-based equivalencies to spatial decomposition methods like the Hirshfeld or Bader partitioning. The goal of this project is to implement these approaches in the OrbsTools module of ChemTools (). | ||
| Line 347: | Line 359: | ||
'''Mentors:''' Gabriela Sánchez Díaz (sanchezg at mcmaster dot ca), Marco Martinez-Gonzalez (mmg870630 at gmail dot com), Paul Ayers (ayers at mcmaster dot ca), and Farnaz Heidar-Zadeh (farnaz.heidarzadeh at queensu dot ca). | '''Mentors:''' Gabriela Sánchez Díaz (sanchezg at mcmaster dot ca), Marco Martinez-Gonzalez (mmg870630 at gmail dot com), Paul Ayers (ayers at mcmaster dot ca), and Farnaz Heidar-Zadeh (farnaz.heidarzadeh at queensu dot ca). | ||
=== Project: [350 hours] Model Hamiltonian (interface to FCIDump and other solid-state codes) === | |||
'''Brief Explanation: '''In many cases, the true many-electron Hamiltonian is intractable to solve, so model Hamiltonians that capture key features of the physicochemical system qualitatively. Classic examples include the (extended) Hubbard, Ising, and Heisenberg model Hamiltonians. The goal of the package is to build an framework for constructing Model Hamiltonians and outputting them into a format that is conducive to traditional packages for solving the quantum many-body problem. | '''Brief Explanation: '''In many cases, the true many-electron Hamiltonian is intractable to solve, so model Hamiltonians that capture key features of the physicochemical system qualitatively. Classic examples include the (extended) Hubbard, Ising, and Heisenberg model Hamiltonians. The goal of the package is to build an framework for constructing Model Hamiltonians and outputting them into a format that is conducive to traditional packages for solving the quantum many-body problem. | ||
| Line 500: | Line 512: | ||
These ideas would likely benefit two or more projects. | These ideas would likely benefit two or more projects. | ||
===Project [350 hours]: Computational Chemistry Repository Server === | |||
'''Brief explanation:''' Providing a web server to store and organize computational chemistry results is extremely needed. Previous work includes MonogoChemServer (https://github.com/OpenChemistry/mongochemserver) and OpenChemVault (https://github.com/cclib/openchemvault). The former includes visualizations for vibrations, spectra, etc. but is based on the out-dated Girder data management platform. The latter handles its own data management, but doesn't offer the same range of visualizations. | |||
'''Expected results:''' An extensible, maintainable server framework built on FastAPI and MongoDB, which can use cclib to import computational files from any source, provide basic charts and visualizations (e.g., potential energy curves, spectra, vibrations). | |||
Assuming the project is successful, a manuscript will be prepared for publication in an appropriate journal. | |||
'''Prerequisites:''' Experience with Python and Flask and at least some understanding of computational chemistry. Some understanding with FastAPI would be welcome. | |||
'''Mentor:''' Geoffrey Hutchison (geoffh at pitt.edu) and Marcus Hanwell | |||
===Project [350 hours]: Active Learning Bayesian Geometry Optimizer === | |||
'''Brief explanation:''' Bayesian Optimization (e.g., BoTorch) is a highly effective method for global optimization of "black box" functions. Since it does not require gradients, it can efficiently optimize ML models or even electronic structure methods without derivatives (e.g., coupled-cluster). | |||
The Hutchison group and others have shown global optimization of conformers using a Bayesian approach. | |||
Instead, the project will focus on '''local''' optimization in internal coordinates (e.g., bond lengths, angles, dihedrals) using an active learning approach. In other words, after optimizing a C-C single bond, the optimizer will know the functional form (e.g., Morse potential) and bond length / angles for similar motifs. | |||
'''Expected results:''' A PyTorch-based optimizer with support for several computational models (e.g., force fields, quantum chemistry, ML) and benchmarking comparisons between naive optimization and few-shot learning. | |||
Assuming the project is successful, a manuscript will be prepared for publication in an appropriate journal. | |||
'''Prerequisites:''' Experience with Python and PyTorch and at least some understanding of Bayesian optimization and computational chemistry. | |||
'''Mentor:''' Geoffrey Hutchison (geoffh at pitt.edu) | |||
Latest revision as of 18:09, 29 March 2023
Guidelines
Open Chemistry is an umbrella for projects in chemistry, materials science, biochemistry, and related areas.
One important factor is that GSoC in 2023 will include both shorter projects (~175 hours) and longer projects (~350 hours). You should consider the appropriate timeline for your project proposal. We have indicated in the project totals where we suggest particular lengths.
Contributors can also decide on the number of weeks (e.g., spreading the project time over multiple weeks).
If you are unsure of the scope of a project, please reach out and discuss BEFORE the proposal deadline (20 March to 4 April).
When possible, submitting drafts a week or more in dance of the proposal deadline is preferred because we can make suggestions towards your proposal.
We have gathered a pool of interested mentors together who are seasoned developers in each of these projects. We welcome original ideas in addition to what's listed here - please suggest something interesting for open source chemistry!
Adding Ideas
When adding a new idea to this page, please try to include the following information:
- Size of the project (~175 hours of work) or (~350 hours of work)
- A brief explanation of the idea.
- Expected results/feature additions.
- Any prerequisites for working on the project.
- Links to any further information, discussions, bug reports etc.
- Any special mailing lists if not the standard mailing list for the project
- Your name and email address for contact (if willing to mentor, or nominated mentor).
Proposal Guidelines
Students need to write and submit a proposal, we have added the applying to GSoC page to help guide our students on what we would like to see in those proposals.
Avogadro 2 Project Ideas
Avogadro 2 is a chemical editor and visualization application, it is also a set of reusable software libraries written in C++ using principles of modularity for maximum reuse. We offer permissively licensed, open source, cross platform software components in the Avogadro 2 libraries, along with an end-user application with full source code, and binaries.
Project [350 hours]: Scripting Bindings
Brief explanation: Implement an embedded scripting language (i.e., Python) in Avogadro 2
Expected results: Enable an embedded scripting console as well as support for implementing modular extensions (tools, rendering, etc.) in Python. Python bindings exist, using PyBind11 with the new codebase, and the Avogadro 2 core libraries are pip installable. Extending the coverage of the API from the rudimentary parts of core/io would be a good starting point. An ideal solution would connect to PySide, to allow scripting to add UI like menu items, windows, etc. and provide documentation and example scripts. The interface should be maintainable as new classes and methods are added.
Example scripts, documentation, are highly encouraged.
Prerequisites: Experience in C++ and Python, some experience with PyBind11, Qt for Python, PySide suggested.
Mentor: Geoff Hutchison (geoffh at pitt dot edu) or Marcus D. Hanwell (mhanwell at bnl dot gov)
Project [175 or 350 hours]: Tools for Interactive Molecular Dynamics
Brief explanation: Building solvent boxes, implementing standard molecular dynamics using in-progress optimization framework. The scope could be 175 or 350 hours - please discuss what scale project you have in mind.
Expected results: Avogadro (v1) has interactive force field optimization allowing building and manipulation (e.g., push-pull atoms into position). Some users call this 'video game mode' ;-) A new optimization framework is in progress, including calling external programs for energies and forces. The project would enable building out MD simulations, including tools to add water or solvent boxes, build larger systems (e.g., via PackMol integration) and implement simple MD integration and thermostats.
'Prerequisites: Experience in C++, ideally with knowledge of molecular dynamics methods and tools. Some Python would be helpful
Mentor: Geoff Hutchison (geoffh at pitt dot edu)
Project [350 hours]: Improved Rendering with Geometry Shaders
Brief explanation: Our current rendering code needs updating to the OpenGL Core Profile and Optimization with Geometry Shaders
Expected results: An efficient GPU-enabled surface generation and rendering framework using geometry shaders to provide dynamic level of detail, improved depth-of-focus and rendering quality.
'Prerequisites: Experience in C++, ideally with knowledge of OpenGL shaders. Some understanding of quantum chemistry would be helpful.
Mentor: Geoff Hutchison (geoffh at pitt dot edu)
Project [175 or 350 hours]: Improved Selection and/or Molecular Find-and-Replace
Brief explanation: Improved support for selection tools (e.g., "select all water molecules" and "select everything within X Å of my mouse-click") and tools for bulk editing of structures using a "find and replace" interface (e.g., replace 10% of all gold atoms with silver or "change this molecular pattern to this new functional group")
Expected results: A set of new selection tools and commands and/or a set of tools to substitute atoms and molecular fragments in the builder. For example, the Wilmer group published "MOFUN" (https://doi.org/10.1039/D2DD00044J) a package to replace specific fragments, and many packages such as Open Babel and RDKit support SMIRKS or reaction SMILES rules. Ideas to improve the usability are highly welcome.
'Prerequisites: Experience in C++, some experience with Python will be helpful.
Mentor: Geoff Hutchison (geoffh at pitt dot edu)
Project [175 hours]: Integrate with RDKit
Brief explanation: Integrate the RDKit toolkit into Avogadro for conformer sampling and force field optimization
Expected results: RDKit is a BSD-licensed cheminformatics toolkit with a wide range of features useful for Avogadro 2. Most notably, RDKit offers efficient and accurate 3D coordinate generation, conformer sampling, and force field optimization. Implement a connection between Avogadro objects (molecules and atoms) and RDKit objects and implement conformer sampling and force field optimization code.
'Prerequisites: Experience in C++, some experience with Python will be helpful.
Mentor: Geoff Hutchison (geoffh at pitt dot edu)
Open Babel Project Ideas
Open Babel is an open toolbox for chemistry, designed to speak the many languages of chemical data. It's an open, collaborative project allowing anyone to search, convert, analyze, or store data from molecular modeling, chemistry, solid-state materials, biochemistry, or related areas.
Project [175 hours]: Integrate CoordGen library
Expected results: Schrodinger has released a BSD-licensed library for 2D chemical structure layout (https://github.com/schrodinger/coordgenlibs) and it has been successfully integrated into RDKit. The student will be responsible for integrating CoordGen into Open Babel. Code will be written in C++.
Mentor: Geoff Hutchison (geoffh at pitt dot edu)
Project [175 hours]: Implement MMTF format
Brief explanation: Implementation of MMTF file format in OpenBabel.
Expected results:' Macromolecular Transmission Format (MMTF) is a new compact binary format to transmit and store biomolecular structural data quickly and accurately (http://mmtf.rcsb.org). Your task is to implement support for this format in the OpenBabel open-source cheminformatics toolkit (http://openbabel.org). Code will be written in C++.
Mentor: Geoff Hutchison (geoffh at pitt dot edu) or David Koes (dkoes at pitt dot edu)
Project [175 hours]: Test Framework Overhaul
Brief explanation: Automated testing is an important part of maintaining code quality. This project will improve the current testing regime of openbabel.
Expected results: A comprehensive test framework that automates the generation of unit tests for all supported languages and simplifies the creation of new test cases will be implemented. The student will be responsible for choosing the most appropriate framework, porting existing test cases, and expanding the test suite to enhance code coverage.
Prerequisites: Experience in C++. Knowledge of modern software engineering practices or test frameworks is ideal.
Mentor: Geoff Hutchison (geoffh at pitt dot edu), David Koes (dkoes at pitt dot edu), the OpenBabel development community.
Project [350 hours]: Develop a JavaScript version of Open Babel
Brief explanation: Building on existing work, you will use Emscripten to compile the C++ codebase of Open Babel to JavaScript. This will make it easy to write in-browser applications that need cheminformatics functionality.
Expected results: Following from work described in a recent paper (https://pubs.acs.org/doi/abs/10.1021/acs.jcim.7b00434), a JavaScript version of the Open Babel toolkit will be created. The generation of any necessary wrappers should be automated to allow it to track changes in the Open Babel API.
Ideally, the project will adapt a core JavaScript library openbabel.js that allows modules, such as file formats to be imported separately (e.g., smilesformat.js, pdbformat.js, xyzformat.js, etc.)
Prerequisities: Some experience in C++, and also with JavaScript.
Mentor: Geoff Hutchison (geoffh at pitt dot edu)
Project [350 hours]: Develop a validation and standardization filter
Brief explanation: Given a particular molecular structure, can we say how chemically plausible is it, and use this as to filter or warn about problems (e.g., undefined stereo centers)?
Expected results: Given a set of reference structures (e.g. ChEMBL), it should be possible to build a model that can say how normal/unusual a query structure is. For example, given a set of drug-like molecules, a molecule with a ruthenium atom might be considered unusual; or given any set of molecules, a 5-coordinate carbon is unusual.
Such a model could be used as a filter, or as a warning to flag up problematic structures.
To be clear "model" does not necessarily mean a machine learning model, instead as a set of filters. (An ML-based model for tautomers might be useful, however.)
Code could be modeled on MolVS using RDKit [[1]]
Prerequisites: Experience in C++ or Python, and an interest in data science or statistics.
Mentor: Geoff Hutchison (geoffh at pitt dot edu)
cclib Project Ideas
cclib is an open source library, written in Python, for parsing and interpreting the results of computational chemistry packages. The goals of cclib are centered around the reuse of data obtained from these programs when stored in program-specific output files.
Project: [175 or 350 hours] Implement new parsers
Brief explanation: There are outstanding issues on GitHub for supporting more programs (e.g. CFOUR, xtb, NBO, GAMESS dat, MRCC, DIRAC), and parsing binary files for various QM programs (e.g. Gaussian, NWChem, and ORCA). There may also be more programs missing that haven't been considered.
Expected results: Implement parsers for one or more new programs/formats, generate test data, and write unit and regression tests for each parser.
Prerequisites: Experience with Python, basic familiarity with computational chemistry programs, and access to the program(s) needed to generate the test data.
Mentors: Eric Berquist (eric.john.berquist at gmail dot com) and/or Shiv Upadhyay (shivnupadhyay at gmail dot com) and/or Adam Tenderholt (atenderholt at gmail dot com) and/or Karol Langner (karol.langner at gmail dot com)
Project: [175 or 350 hours] Implement new bridges
Brief explanation: There are outstanding issues on GitHub for more integrations with external programs (e.g. chemfiles, RDKit) via their Python bindings. There may also be more programs missing that haven't been considered.
Expected results: Implement bridges for one or more new programs, along with writing unit tests and documentation for each bridge.
Prerequisites: Experience with Python and ideally familiarity with the program that is being bridged.
Mentors: Eric Berquist (eric.john.berquist at gmail dot com) and/or Shiv Upadhyay (shivnupadhyay at gmail dot com) and/or Adam Tenderholt (atenderholt at gmail dot com) and/or Karol Langner (karol.langner at gmail dot com)
Project [350 hours]: Implement new methods
Brief explanation: There are outstanding issues on GitHub for more analysis methods being added directly to cclib (e.g. calculating geometric parameters). There may also be other methods that are desirable to include which haven't been considered.
Expected results: Implement one or more new methods, along with writing unit tests and documentation for each method.
Prerequisites: Experience with Python and familiarity with the method(s) being added, depending on the complexity of the method.
Mentors: Eric Berquist (eric.john.berquist at gmail dot com) and/or Shiv Upadhyay (shivnupadhyay at gmail dot com) and/or Adam Tenderholt (atenderholt at gmail dot com) and/or Karol Langner (karol.langner at gmail dot com)
Project [350 hours]: Julia bindings
Brief explanation: The Julia programming language (https://julialang.org/) is growing in popularity for computational chemistry as a language that both production-level computation and analysis can be performed in seamlessly. In order to analyze computational chemistry outputs from traditional programs in Julia, rather than reimplement all cclib functionality in Julia, we should be able to call cclib from Julia directly and reuse its core functionality.
Expected results: Julia bindings to cclib IO functionality and a Julia-native representation of cclib data objects, with each cclib attribute accessible as a native Julia type. The bindings should be available on the default Julia package registry. The remainder of the project is more open-ended, but an example application of using the bindings would be ideal.
Prerequisites: Experience with Python and/or Julia, and ideally some familiarity with important quantities from computational chemistry outputs.
Mentors: Eric Berquist (eric.john.berquist at gmail dot com)
Project [350 hours]: Additional visualization for OpenChemVault
Brief explanation: OpenChemVault (https://github.com/cclib/openchemvault) is capable of parsing output files, storing them, and displaying geometries, but any sort of additional visualization (such as plotting molecular orbitals or spectra) is missing. The capabilities of GaussSum (http://gausssum.sourceforge.net/) are a possible starting point.
Expected results: Implement one or more new visualizations for the OpenChemVault web interface.
Prerequisites: Experience with Python common visualizations that are desirable for computational chemistry outputs. No previous experience with JavaScript is necessary.
Mentors: Eric Berquist (eric.john.berquist at gmail dot com) and/or Shiv Upadhyay (shivnupadhyay at gmail dot com) and/or Adam Tenderholt (atenderholt at gmail dot com) and/or Karol Langner (karol.langner at gmail dot com)
RDKit Project Ideas
The RDKit is a BSD licensed open source cheminformatics toolkit written in C++ with wrappers for use from Python, Java, C#, and JavaScript. The RDKit also provides "cartridge" functionality that allows chemical searching in the open-source relational database PostgreSQL.
Project [350 hours]: Implement Molecular Interaction Fields calculations in the RDKit
Brief explanation: There is an old PR for the RDKit that implements molecular interaction fields: https://github.com/rdkit/rdkit/pull/318. This was never merged because the author ran out of time. At this point a lot of work would be required to update and finish this PR, but the results would be super useful for the RDKit community.
Expected results: A C++ implementation of the GRID calculator code along with a robust set of test cases. Wrappers for the calculator so that it is accessible from within the Python and SWIG (Java and C#) wrappers.
Prerequisites: C++
Mentor: Greg Landrum (greg.landrum at t5informatics dot com)
Project [175 or 350 hours]: Implement additional fingerprints in the RDKit
Brief explanation: There are a number of chemical fingerprint types which it would be useful to have natively available in the RDKit; in this project you will implement one or more of them. Some ideas for fingerprints to be included are:
- Pubchem fingerprint: https://ftp.ncbi.nlm.nih.gov/pubchem/specifications/pubchem_fingerprints.pdf
- CSFP: https://doi.org/10.1021/acs.jcim.9b00571
- Physicochemical property fingerprints: porting the existing Python implementation (https://github.com/rdkit/rdkit/blob/7153918af4dff37c768577441c5286b425e6bf3d/rdkit/Chem/AtomPairs/Sheridan.py) to C++
The number to be implemented depends on whether you are doing this as a 175 or 350 hour project.
Expected results: A C++ implementation of the new fingerprints along with a robust set of test cases. Wrappers for the calculators so that they are accessible from with the Python and SWIG (Java and C#) wrappers.
Prerequisites: C++ and some Python
Mentor: Greg Landrum (greg.landrum at t5informatics dot com)
QC-Devs Project Ideas
QC-Devs (https://qcdevs.org/) develops various free, open-source, and cross-platform libraries for scientific computing, especially theoretical and computational chemistry. Our goal is to make programming accessible to chemists and promote precepts of sustainable software development. The two main pieces of the QC-Devs ecosystem are:
HORTON (electronic structure theory): https://quantumelephant.org/
ChemTools (molecular structure and reactivity): https://chemtools.org/
All our repositories are hosted on Theochem organization (https://github.com/theochem) on GitHub.
Project [175 hours or 350 hours]: Visualization of Molecular Structure and Reactivity
Brief Explanation: ChemTools (https://github.com/theochem/chemtools) is a post-processing library for extracting chemical insight from quantum chemistry calculations. Currently, ChemTools relies on Visual Molecular Dynamics (VMD) and Matplotlib for visualization. ChemTools has the functionality to generate visualization scripts for VMD, so the user can easily generate informative plots like iso-surface of electron density colored by electrostatic potential. Visualization of (annotated) molecular structures and molecular structure changes along reaction pathways are also of interest, but the implementations are unpolished.
Expected Results:
175 hours: Add functionality to ChemTools to generate visualization scripts for ChimeraX (https://www.cgl.ucsf.edu/chimerax/). The current functionality for VMD can be used as a template.
350 hours: Add ChemTools as a back-end for SEQCROW (https://cxtoolshed.rbvi.ucsf.edu/apps/seqcrow), a free and open-source bundle (https://github.com/QChASM/SEQCROW) in the ChimeraX toolshed (https://cxtoolshed.rbvi.ucsf.edu/) for building molecules and interacting with the output of quantum chemistry calculations.
Difficulty Level: Intermediate (175) to High-Intermediate (350)
Relevant Skills: Experience with Python, visualization, and software interfacing (350)
Mentors: Ali Tehrani (alirezatehrani24 at gmail dot com), Gabriela Sanchez Diaz (sanchezg at mcmaster dot ca), Farnaz Heidar-Zadeh (farnaz.heidarzadeh at queensu dot ca), and Esteban Vohringer-Martinez (estebanvohringer at qcmmlab dot com).
Project [175 or 350 hours]: Extended interoperability of GOpt and Quantum Chemistry Software
Brief Explanation: ChemTools (https://github.com/theochem/chemtools) is a post-processing library for extracting chemical insight from quantum chemistry calculations. Currently, ChemTools relies on modules of the HORTON library to compute the basic quantities required for its analysis. The goal of this project is to extend the interoperability of ChemTools, so that it can use the Psi4 (https://github.com/psi4) & PySCF (https://github.com/pyscf/pyscf) packages and take advantage of their features.
Expected Results:
175 hours: Writing wrappers for Psi4 or PySCF to compute various quantum mechanical properties and provide those properties to ChemTools for further analysis. The current wrappers for HORTON can be used as a template. Both Psi4 & PySCF have Python interfaces.
350 hours: Writing wrappers for Psi4 and PySCF.
Difficulty Level: Intermediate
Relevant Skills: Experience with scientific Python, advanced Numpy, object-oriented programming, and knowledge of quantum chemistry software
Mentors: Gabriela Sanchez Diaz (sanchezg at mcmaster dot ca), Ali Tehrani (alirezatehrani24 at gmail dot com), and Farnaz Heidar-Zadeh (farnaz.heidarzadeh at queensu dot ca).
Project [175 hours] Molecule Alignment with Procrustes Algorithm
Brief Explanation: Procrustes (https://github.com/theochem/procrustes) is a library for finding the optimal transformation that makes two matrices as close as possible to each other. Permutation Procrustes methods can be used for molecular alignment (J Math Chem (2013) 51:927ΓÇô936). The goal of this progress is to develop a utility that uses the Procrustes package to perform molecular alignment.
Expected Results: An open-source Python software.
175 hours: Using the Procrustes package, write a utility that takes two molecular structures and optimizes their alignment. In addition to simply optimizing the structural alignment, provide atom-atom mapping and extensions to more general problems (e.g., multi-molecule alignment).
Difficulty Level: Advanced
Relevant Skills: Experience with scientific Python, advanced Numpy, and object-oriented programming.
Mentors: Fanwang Meng (fwmeng88 at gmail dot com) and Paul Ayers (ayers at mcmaster dot ca).
Project [350 hours]: Faster Molecular Integrals with Density-Fitting
Brief Explanation: GBasis (https://github.com/theochem/gbasis) is a library for evaluating and analytically integrating Gaussian-type orbitals and their related quantities, especially molecular integrals. In many applications, the computational bottleneck is the evaluation of two-electron integrals, as the number of two-electron integrals grows as the fourth power of the basis-set size. By introducing an auxiliary, density-fitting, basis, this power is reduced to the third power of the basis-set size, which in many cases eliminates the computational bottleneck, since there are often other facets of the computation that scale more severely than this. The goal of this project is to implement density-fitting methods into GBasis.
Expected Results:
350 hours: Extension of GBasis to support density fitting. This involves expanding products of basis functions in the auxiliary basis, evaluating 2-electron integrals in the auxiliary basis, and using these two entities to construct molecular integrals more efficiently.
Difficulty Level: Intermediate to Advanced
Relevant Skills: Experience with scientific Python, advanced Numpy, and object-oriented programming.
Mentors: Ali Tehrani (alirezatehrani24 at gmail dot com), Gabriela Sanchez Diaz (sanchezg at mcmaster dot ca), and Paul Ayers (ayers at mcmaster dot ca).
Project [350 hours]: Computing The Pair Density From Wave-function
Brief Explanation: The electron pair density represents the probability of observing two electrons at two points in space. It provides key quantitative and qualitative information about electron correlation, as well as qualitative information about chemical bonding and, in particular, about how Lewis structures emerge from quantum mechanics.
Expected Results:
350 hours: To provide a Python function to compute the pair-density using GBasis (https://github.com/theochem/gbasis) as a Python function, starting from wave-function information that is read with IOData (https://github.com/theochem/iodata). Key indicators like the intracule and extracule should be supported.
Difficulty Level: Intermediate
Relevant Skills: Experience with scientific Python, advanced Numpy, and object-oriented programming.
Mentors: Ali Tehrani (alirezatehrani24 at gmail dot com), Gabriela Sanchez Diaz (sanchezg at mcmaster dot ca), and Paul Ayers (ayers at mcmaster dot ca).
Project: [350 hours] Extended Interoperability of Denspart to Open Force Field Software
Brief Explanation: Partitioning a chemical system and its various properties into atomic contributions is not only qualitatively useful for chemical analysis but important for quantitative computational chemistry (e.g., charges in molecular mechanics force fields). This can be done by partitioning the molecular electron density into fuzzy atomic densities using Denspart ( package and class within ChemTools package), but there are alternative approaches based on topological partitioning of molecular density or partitioning of molecular orbitals. The goal of this project is to develop a unified framework determining the molecular contributions of specified atoms and/or functional groups.
Expected Results:
- Design a unified DensPart API with interfaces to the two aforementioned packages.
- Develop a module for computing various non-covalent force-field energy terms, especially those that are useful for molecular mechanics force field parameterization/development.
- Design a framework for enhancing molecular modeling using Denspart (based on quantum chemistry calculations)
Difficulty Level: Intermediate
Relevant Skills: Experience with scientific Python, advanced Numpy, and object-oriented programming, familiarity with quantum chemistry, molecular mechanics, and force-fields.
Mentors: Farnaz Heidar-Zadeh (farnaz.heidarzadeh at queensu dot ca), Toon Verstraelen (Toon.Verstraelen at ugent dot be), and Esteban Vohringer-Martinez (estebanvohringer at qcmmlab dot com).
Project: [350 hours] Extended Interoperability of IOData to Open Force Field Software
Brief Explanation: IOData () is a free and open-source Python library for parsing, storing, and converting various file formats commonly used by quantum chemistry, molecular dynamics, and plane-wave density-functional-theory software programs. It also supports a flexible framework for generating input files for various software packages. The goal of this project is to extend the IOData framework to describe a chemical system for molecule mechanics (MM) modeling, specifically in the context of non-covalent interactions of open force-fields.
Expected Results:
- Embed a compatible data structure for non-bonded parameters derived from electron density partitioning methods into the IOData framework.
- Generate an Open Force Field Toolkit compatible output ()
- Extend IOData framework to accommodate bonded parameters to have a full interface for molecular modeling.
Difficulty Level: Intermediate
Relevant Skills: Experience with scientific Python, advanced Numpy, and object-oriented programming, familiarity with molecular mechanics and force-fields.
Mentors: Farnaz Heidar-Zadeh (farnaz.heidarzadeh at queensu dot ca), Toon Verstraelen (Toon.Verstraelen at ugent dot be), Esteban Vohringer-Martinez (estebanvohringer at qcmmlab dot com).
Project: [350 hours] Deriving Atomic and Bond Properties from Molecule Orbital Partitioning
Brief Explanation: In traditional orbital-based atomic partitioning methods for molecular wavefunctions, one observes that the results are strongly dependent on the molecular basis set, and tend to work poorly for large atom-centered basis sets and be inapplicable to delocalized (e.g., plane wave) basis functions. One way to avoid these issues is using various sorts of quasi-atomic orbitals (QUAOs), including quasi-atomic molecular basis orbitals (QUAMBO), intrinsic atomic orbitals (IAOs), and other choices. Another alternative is to use (approximate) orbital-based equivalencies to spatial decomposition methods like the Hirshfeld or Bader partitioning. The goal of this project is to implement these approaches in the OrbsTools module of ChemTools ().
Expected Results:
- Design an API for various quasi-atomic orbital methods and implement QUAMBO, IAO, QUAO and possibly other quasi-atomic orbital methods.
- Reenvision the Mulliken and Lowdin atom/bond analysis based thereupon.
- Implement the orbital-based partitioning based on an arbitrary spatial partitioning method, the Mulliken and Lowdin atom/bond analysis based thereupon.
Difficulty Level: Intermediate to Hard
Relevant Skills: Experience with scientific Python, advanced Numpy, and object-oriented programming, familiarity with quantum chemistry methods.
Mentors: Gabriela Sánchez Díaz (sanchezg at mcmaster dot ca), Marco Martinez-Gonzalez (mmg870630 at gmail dot com), Paul Ayers (ayers at mcmaster dot ca), and Farnaz Heidar-Zadeh (farnaz.heidarzadeh at queensu dot ca).
Project: [350 hours] Model Hamiltonian (interface to FCIDump and other solid-state codes)
Brief Explanation: In many cases, the true many-electron Hamiltonian is intractable to solve, so model Hamiltonians that capture key features of the physicochemical system qualitatively. Classic examples include the (extended) Hubbard, Ising, and Heisenberg model Hamiltonians. The goal of the package is to build an framework for constructing Model Hamiltonians and outputting them into a format that is conducive to traditional packages for solving the quantum many-body problem.
Expected Results:
1. Provide support for Huckel/Hubbard/PPP parameters for various atom and connectivity types.
2. Write user-friendly APIs for constructing model Hamiltonians of various types: Heisenberg, Ising, generalized Richardson-Gaudin, t-J, and t-J-U-V models.
3. Write utilities to output model Hamiltonians into formats conducive to external programs, including FCIDump and Triqs.
4. Create and write example tutorials.
Difficulty Level: Intermediate to Hard
Relevant Skills: Experience with scientific Python, advanced Numpy, and object-oriented programming, and familiarity with quantum mechanics.
Mentors: Gabriela Sánchez Díaz (sanchezg at mcmaster dot ca), Marco Martinez-Gonzalez (mmg870630 at gmail dot com), Paul Ayers (ayers at mcmaster dot ca), and Farnaz Heidar-Zadeh (farnaz.heidarzadeh at queensu dot ca).
CalcUS Project Ideas
CalcUS is a platform aiming to democratize access to quantum chemistry by providing a user-friendly web-based interface to simplify running and analyzing quantum mechanical calculations.
Project [175 hours]: Develop large-scale calculation management tools
Brief Explanation: Quantum chemistry projects can involve performing calculations on a large number of structures (10-100) with different parameters. CalcUS should have features to make this process seamless and highly automated, from launching the calculations to reporting the results.
Expected Results: Create a variation of the calculation web UI, aimed specifically at batch calculations with variable parameters, design and implement the workflow to handle these batch calculations, implement results gathering and reporting in a convenient format, write relevant unit and/or integration tests.
Prerequisites: Knowledge of HTML, Javascript and Python. Familiarity with JQuery, Django and PostgreSQL is helpful.
Mentor: Raphaël Robidas (raphael dot robidas at usherbrooke dot ca) Claude Legault (claude dot legault at usherbrooke dot ca)
ccinput Project Ideas
ccinput is a library and standalone tool to create computational chemistry input files.
Project [350 hours]: Add support for NWChem
Brief Explanation: Implementing the creation of NWChem input files for most of its features.
Expected Results: Implementing the creation of NWChem input files which follow the correct structures, implementing support of core keywords and modifiers, adding the necessary static data about NWChem (supported methods, solvents, etc.), creation of extensive unit tests for all new features, writing any necessary documentation about these new features.
Prerequisites: Knowledge of Python. Familiarity with quantum chemistry is helpful, but not required.
Mentor: Raphaël Robidas (raphael dot robidas at usherbrooke dot ca), Claude Legault (claude dot legault at usherbrooke dot ca)
3Dmol.js Project Ideas
3Dmol.js is a modern, object-oriented JavaScript library for visualizing molecular data that is forked from GLmol. A particular emphasis is placed on performance.
Project [175 hours]: More cartoon options for nucleic acids.
Brief explanation: Implement additional visualizations of nucleic acids.
Expected results: See https://github.com/3dmol/3Dmol.js/issues/559
Prerequisites: Experience with JavaScript and client-server programming, some experience with OpenGL/WebGL ideal, but not necessary.
Mentor: David Koes (dkoes@pitt.edu)
Project [175 or 350 hours]: Improve 3Dmol.js
Brief explanation: Make significant improvements to 3Dmol.js functionality or performance.
Expected results: This is an open-ended project that must be driven by the applicant. A strong proposal will identify significant shortcomings in the current code and explain how it will be addressed. The GitHub Issues page may provide some ideas. A proposal must include a significant initial pull request.
Prerequisites: Experience with JavaScript and client-server programming, some experience with OpenGL/WebGL ideal, but not necessary.
Mentor: David Koes (dkoes@pitt.edu)
gnina Project Ideas
gnina is a C/C++ framework for applying deep learning to molecular docking.
Project [175 or 350 hours]: Improve gnina
Brief explanation: Make significant improvements to gnina functionality or performance.
Expected results: This is an open-ended project that must be driven by the applicant. A strong proposal will identify significant shortcomings in the current code and explain how it will be addressed. The GitHub Issues page may provide some ideas. A proposal must include a significant initial pull request.
Prerequisites: Experience with CUDA/C/C++ programming and the basics of deep learning.
Mentor: David Koes l (dkoes@pitt.edu)
DeepChem Project Ideas
DeepChem aims to provide a high quality open-source toolchain that democratizes the use of deep-learning in drug discovery, materials science, quantum chemistry, and biology. Additional project ideas are discussed at https://forum.deepchem.io/
Project [350 hours]: Layer Documentation
Brief explanation: DeepChem is moving towards a concept of first class layers. Improving the documentation for existing layers will help us make our current collection of layers more useful for the community.
Expected results: This project should also add a tutorial for using the layers to the DeepChem tutorial series, and should plan to add a few new layers as well.
Prerequisites: PyTorch/TensorFlow, Python
Mentor: Bharath Ramsundar (bharath at deepforestsci dot com)
Project [350 hours]: PyTorch Porting
Brief explanation: DeepChem is shifting towards using PyTorch as its primary backend, but many models are still implemented in TensorFlow. A good project could be to pick a TensorFlow model or two, then port its layers and model into PyTorch along with suitable unit tests.
Expected results: At least one model should be ported from TensorFlow to PyTorch successfully with associated unit tests. See See https://github.com/deepchem/deepchem/issues/2863
Prerequisites: PyTorch/TensorFlow, Python
Mentor: Bharath Ramsundar (bharath at deepforestsci dot com)
Project [350 hours]: HuggingFace Integration
Brief explanation: HuggingFace Integration: Last year, we had a few student projects explore HuggingFace/DeepChem integration, but these projects were not able to merge in HuggingFace models into DeepChem.
Expected results: This project would create a working HuggingFace model in DeepChem along with tutorials on how to use HuggingFace with DeepChem.
Prerequisites: PyTorch/TensorFlow, Python
Mentor: Bharath Ramsundar (bharath at deepforestsci dot com)
Project [350 hours]: Improved PINNs Support
Brief explanation: Improving our PINNs Support: One of the exciting new features in DeepChem 2.6.0 is support for PINNs, a class of techniques to solve PDEs with neural networks. The API for this class is still rudimentary and supports only a limited class of models and requires handcoding the loss.
Expected results: Extend the API to allow for a broader class of PDEs to be implemented. I’d suggest using Schrodinger’s equation as a test since Schrodinger can be solved in 1D as a toy and extended to arbitrarily high dimensions for larger molecules.
Prerequisites: PyTorch/TensorFlow, Python
Mentor: Bharath Ramsundar (bharath at deepforestsci dot com)
Project [350 hours]: Improved Equivariance Support
Brief explanation: Improve Equivariant Support: DeepChem has no support for equivariant models. Given the increasing importance of equivariance for scientific machine learning this is a major oversight.
Expected results: This project would aim to add a tutorial about equivariant modeling and add an equivariant model to DeepChem. You may want to use e3nn or another library to facilitate implementation.
Prerequisites: PyTorch/TensorFlow, Python
Mentor: Bharath Ramsundar (bharath at deepforestsci dot com)
Project [350 hours]: Improved Antibody Support
Brief explanation: Improving Antibody Support: DeepChem at present doesn’t have much tooling or support for working with anbtibodies.
Expected results: This project would add suitable antibody datasets to MoleculeNet and create a tutorial walking users through antibody design and modeling with DeepChem. If necessary, students may add antibody-specific models as well.
Prerequisites: PyTorch/TensorFlow, Python
Mentor: Bharath Ramsundar (bharath at deepforestsci dot com)
Miscellaneous Project Ideas
These ideas would likely benefit two or more projects.
Project [350 hours]: Computational Chemistry Repository Server
Brief explanation: Providing a web server to store and organize computational chemistry results is extremely needed. Previous work includes MonogoChemServer (https://github.com/OpenChemistry/mongochemserver) and OpenChemVault (https://github.com/cclib/openchemvault). The former includes visualizations for vibrations, spectra, etc. but is based on the out-dated Girder data management platform. The latter handles its own data management, but doesn't offer the same range of visualizations.
Expected results: An extensible, maintainable server framework built on FastAPI and MongoDB, which can use cclib to import computational files from any source, provide basic charts and visualizations (e.g., potential energy curves, spectra, vibrations).
Assuming the project is successful, a manuscript will be prepared for publication in an appropriate journal.
Prerequisites: Experience with Python and Flask and at least some understanding of computational chemistry. Some understanding with FastAPI would be welcome.
Mentor: Geoffrey Hutchison (geoffh at pitt.edu) and Marcus Hanwell
Project [350 hours]: Active Learning Bayesian Geometry Optimizer
Brief explanation: Bayesian Optimization (e.g., BoTorch) is a highly effective method for global optimization of "black box" functions. Since it does not require gradients, it can efficiently optimize ML models or even electronic structure methods without derivatives (e.g., coupled-cluster).
The Hutchison group and others have shown global optimization of conformers using a Bayesian approach.
Instead, the project will focus on local optimization in internal coordinates (e.g., bond lengths, angles, dihedrals) using an active learning approach. In other words, after optimizing a C-C single bond, the optimizer will know the functional form (e.g., Morse potential) and bond length / angles for similar motifs.
Expected results: A PyTorch-based optimizer with support for several computational models (e.g., force fields, quantum chemistry, ML) and benchmarking comparisons between naive optimization and few-shot learning.
Assuming the project is successful, a manuscript will be prepared for publication in an appropriate journal.
Prerequisites: Experience with Python and PyTorch and at least some understanding of Bayesian optimization and computational chemistry.
Mentor: Geoffrey Hutchison (geoffh at pitt.edu)
Project [350 hours]: OneMol: Google Docs & YouTube for Molecules
Brief explanation: There is a huge need in the research community for improved collaboration tools on web and desktop. OneMol will provide an open API for collaborating on molecular data that both Avogadro and 3Dmol.js will support as reference implementations. OneMol compliant applications will be able to manipulate and view molecular data in real time so that changes made by one client will be propagated to other clients.
File-sharing is a means for sharing data, but it does not share real-time interactions; each user’s data exists in its own isolated environment. Screen-sharing provides a common viewpoint for all participants, but allowing others to interact with the data requires granting access to the host workstation. This approach is needlessly inefficient for the task of collaborating on molecular data, and this inefficiency introduces scalability issues. For example, a simple rotation necessitates a full screen update when the fundamental change in state was a simple change in viewing angles.
The OneMol framework consists of three main components: a client module, embedded in a molecular viewer; a facilitator module that enforces a consistent viewer state between all the clients; and a storage module that stores the raw molecular data. All three modules may coexist on the same machine within the same application. However, we anticipate a more common modality will be to use a publicly hosted facilitator server, since this simplifies network connectivity in the face of firewalls and network address translation.
Expected results: Prototype web services to allow web and/or desktop collaboration using 3DMol as a viewer, likely integrating with existing storage systems (e.g., MongoChem or PQR).
Prerequisites: Experience with scripting, and web services. Interest and experience with databases like MongoDB or DSpace very helpful.
Mentor: David Koes (dkoes@pitt.edu) or Geoffrey Hutchison (geoffh at pitt.edu)