The amount of GPU memory required by alphafold and openfold increases quadratically with the number of amino acids in the system being modeled. If you are modeling a protein complex longer than 1500 residues or so please contact the engineering team before starting as the project will likely require a larger than normal GPU

Using the API

The “AI Folding” api provides a common interface to AI based protein structure prediction tools. The API currently supports openfold (https://github.com/CyrusBiotechnology/openfold) and alphafold ( https://github.com/deepmind/alphafold#alphafold-output ).

To model a single chain protein with alphafold or openfold

cyrus submit ai-folding input.fasta --mode=monomer --ai-tool=alphafold

cyrus submit ai-folding input.fasta --mode=monomer --ai-tool=openfold

Currently, openfold does not support multichain modeling

To model a prokaryotic multichain protein with alphafold

cyrus submit input.fasta --mode=multimer --ai-tool=alphafold --is-prokaryote

To model a eukarytotic multichain protein with alphafold

cyrus submit input.fasta --mode=multimer --ai-tool=alphafold

API Outputs

The AI Folding API runs a post-processing protocol on the results of the AI tool. The purpose of the post-processing is to minimize the output structure into the rosetta energy function and correct any atomic detail level errors in sidechain positioning. The protocol consists of the following three steps

Generate a molprobity report for the models output from the AI tool
Idealize and relax the models output from the AI tool with rosetta
Generate a molprobity report for the rosetta relaxed models.

Alongside the raw output from the AI tool, the API will produce the following directories:

initial_molprobity_reports – Molprobity reports from step 1 of the post-processing
rosetta_relaxed_models – relaxed models from step 2
final_molprobity_reports -- relaxed models from step 3

An explanation of the alphafold output files can be found here: https://github.com/deepmind/alphafold#alphafold-output