Automated ab initio workflows with jobflow, FireWorks, and atomate

Thumbnail

Event details

Date 17.03.2025 20.03.2025
Location
Category Conferences - Seminars

You can apply to participate and find all the relevant information (speakers, abstracts, program,...) on the event website: https://www.cecam.org/workshop-details/automated-ab-initio-workflows-with-jobflow-fireworks-and-atomate-1276.

Registration is required to attend the full event, take part in the social activities and present a poster at the poster session (if any).  However, the EPFL community is welcome to attend specific lectures without registration if the topic is of interest to their research. Do not hesitate to contact the CECAM Event Manager if you have any question.

Description
Automated ab initio calculations have emerged as a powerful tool for computational materials science. Automated workflows offer many benefits over traditional manual approaches, including:

  • Reproducibility: Automation ensures a consistent calculation procedure for complex properties which often require many computational steps and the linking of multiple software packages.
  • Scalability: High-throughput computations enable wide-scale computational searches (often across tens of thousands of compounds) and the generation of large datasets that are essential for machine-learning.
  • Useability: Users benefit from the experience of domain experts with significant previous expertise calculating the properties of interest through well-tested default values and calculation procedures.
Recent years have seen the development of a number of high-throughput workflow tools for automated ab initio calculations. Atomate2 is a library of over 100 computational materials science workflows and is the infrastructure that powers the Materials Project database. Workflows are written using the jobflow library and can be orchestrated across thousands of compute nodes using the FireWorks software. A central goal of atomate2 is to enable the interoperability of density-functional theory (DFT) packages (VASP, CP2K, ABINIT, and QChem are currently supported), with a common API for inputs and outputs. This has enabled two exciting advances: i) heterogenous workflows that take advantage of the different capabilities of DFT packages; and ii) generalisable workflows that can be written once and executed using multiple DFT codes with minimal configuration. Atomate2 also integrates with many other software packages, including AMSET for electron-phonon and electronic transport calculations, phonopy for finite temperature properties, and Lobsterpy for bonding analysis.
The community of developers is highly international with over 20 research groups spanning the UK, Europe, and the US. The user base is similarly diverse both geographically (covering over 15 countries) and scientifically, with multidisciplinary applications in computational chemistry, physics, materials science, and computer science. Over one thousand questions have been asked by users on the dedicated support forum (https://matsci.org) and associated GitHub repositories. Furthermore, the atomate2 software stack has been adopted by industry stakeholders, such as Umicore who are using it to orchestrate high-throughput computational searches.
The community of users and developers mostly consists of high-throughput specialists with previous expertise in this area. Despite this, there is a growing number of researchers new to the field that are interested in both applying existing workflows and developing completely new ones. It is therefore essential to disseminate the knowledge of these tools to broader audiences across computational materials science, chemistry, and physics.

References
[1] A. Jain, S. Ong, W. Chen, B. Medasani, X. Qu, M. Kocher, M. Brafman, G. Petretto, G. Rignanese, G. Hautier, D. Gunter, K. Persson, Concurrency Computat.: Pract. Exper., 27, 5037-5059 (2015)
[2] K. Mathew, J. Montoya, A. Faghaninia, S. Dwarakanath, M. Aykol, H. Tang, I. Chu, T. Smidt, B. Bocklund, M. Horton, J. Dagdelen, B. Wood, Z. Liu, J. Neaton, S. Ong, K. Persson, A. Jain, Computational Materials Science, 139, 140-152 (2017)
[3] J. George, G. Petretto, A. Naik, M. Esters, A. Jackson, R. Nelson, R. Dronskowski, G. Rignanese, G. Hautier, ChemPlusChem, 87, (2022)
[4] J. George, Trends in Chemistry, 3, 697-699 (2021)
[5] C. Chang, V. Deringer, K. Katti, V. Van Speybroeck, C. Wolverton, Nat. Rev. Mater., 8, 309-313 (2023)

Practical information

  • Informed public
  • Registration required

Organizer

  • Alex Ganose (Imperial College London), Janine George (Federal Intitute for Materials Research and Testing In Germany), Gian-Marco Rignanese (Université Catholique de Louvain)

Contact

  • Aude Merola, CECAM Event and Comunication Manager

Event broadcasted in

Share