AlphaFold & ColabFold
AlphaFold and ColabFold
(run on the HMS O2 cluster)
Alphafold (from DeepMind) & ColabFold are new tools for predicting protein structures. They are available as experimental modules on O2. This means that some features may not work as expected, as the code itself was not designed with standard HPC environments in mind. AlphaFold takes fewer parameters, and uses jackhmmer as an MSA generator instead of mmseqs2, which can make it slower than ColabFold for certain inputs. It may also require more resources to run so we recommend using ColabFold for protein structure prediction.
Townhall Resources
Thank you to those who attended our Townhall on June 8th! The following resources from that event are below:
Townhall slides:
Dr. Sergey Ovchinnikov
• Introduction to AlphaFold & ColabFold and what they’re for, examples of input & output
• Slides for Dr. Ovchinnikov’s presentation
Slides for CCB/RC
• Running AlphaFold/ColabFold on O2 – Alex Truong, Calvin Cox, Roger Vargas
• Slides for CCB/RC Presentation
Dr. Mohammed AlQuraishi
• OpenFold: A trainable implementation of AlphaFold
• Colab
• Github page
Dr. Jason Key
• SBGrid & BioGrids: how to further evaluate predictions from AlphaFold/ColabFold
• Slide for Dr. Key’s Presentation
Access AlphaFold & ColabFold
Papers for AlphaFold
Highly accurate protein structure prediction with AlphaFold
Protein complex prediction with AlphaFold-Multimer
Github repo for AlphaFold
Paper for ColabFold
ColabFold: Making protein folding accessible to all
Github repo for ColabFold
Questions?
- General questions about AlphaFold or ColabFold can be sent to Sunil Poudel
- Technical questions about AlphaFold or ColabFold should be sent to Research Computing