Alphafold

Introduction

Alphafold is a protein structure prediction tool developed by DeepMind (Google). It uses a novel machine learning approach to predict 3D protein structures from primary sequences alone. The source code is available on Github. It has been deployed in all RCAC clusters, supporting both CPU and GPU.

It also relies on a huge database. The full database (~2.2TB) has been downloaded and setup for users.

Protein struction prediction by alphafold is performed in the following steps:

Search the amino acid sequence in uniref90 database by jackhmmer (using CPU)
Search the amino acid sequence in mgnify database by jackhmmer (using CPU)
Search the amino acid sequence in pdb70 database (for monomers) or pdb_seqres database (for multimers) by hhsearch (using CPU)
Search the amino acid sequence in bfd database and uniclust30 (updated to uniref30 since v2.3.0) database by hhblits (using CPU)
Search structure templates in pdb_mmcif database (using CPU)
Search the amino acid sequence in uniprot database (for multimers) by jackhmmer (using CPU)
Predict 3D structure by machine learning (using CPU or GPU)
Structure optimisation with OpenMM (using CPU or GPU)

Versions

2.1.1
2.2.0
2.2.3
2.3.0
2.3.1

Commands

run_alphafold.sh

Module

You can load the modules by:

module load biocontainers
module load alphafold

Usage

The usage of Alphafold on our cluster is very straightford, users can create a flagfile containing the database path information:

run_alphafold.sh --flagfile=full_db.ff --fasta_paths=XX --output_dir=XX ...

Users can check its detaied user guide in its Github.

full_db.ff

Example contents of full_db.ff:

--db_preset=full_dbs
--bfd_database_path=/depot/itap/datasets/alphafold/db/bfd/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt
--data_dir=/depot/itap/datasets/alphafold/db/
--uniref90_database_path=/depot/itap/datasets/alphafold/db/uniref90/uniref90.fasta
--mgnify_database_path=/depot/itap/datasets/alphafold/db/mgnify/mgy_clusters_2018_12.fa
--uniclust30_database_path=/depot/itap/datasets/alphafold/db/uniclust30/uniclust30_2018_08/uniclust30_2018_08
--pdb70_database_path=/depot/itap/datasets/alphafold/db/pdb70/pdb70
--template_mmcif_dir=/depot/itap/datasets/alphafold/db/pdb_mmcif/mmcif_files
--max_template_date=2022-01-29
--obsolete_pdbs_path=/depot/itap/datasets/alphafold/db/pdb_mmcif/obsolete.dat
--hhblits_binary_path=/usr/bin/hhblits
--hhsearch_binary_path=/usr/bin/hhsearch
--jackhmmer_binary_path=/usr/bin/jackhmmer
--kalign_binary_path=/usr/bin/kalign

Note

Since Version v2.2.0, the AlphaFold-Multimer model parameters has been updated. The updated full database is stored in depot/itap/datasets/alphafold/db_20221014. For ACCESS Anvil, the database is stored in /anvil/datasets/alphafold/db_20221014. Users need to update the flagfile using the updated database:

run_alphafold.sh --flagfile=full_db_20221014.ff --fasta_paths=XX --output_dir=XX ...

full_db_20221014.ff (for alphafold v2)

Example contents of full_db_20221014.ff (For ACCESS Anvil, please change depot/itap to anvil):

--db_preset=full_dbs
--bfd_database_path=/depot/itap/datasets/alphafold/db_20221014/bfd/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt
--data_dir=/depot/itap/datasets/alphafold/db_20221014/
--uniref90_database_path=/depot/itap/datasets/alphafold/db_20221014/uniref90/uniref90.fasta
--mgnify_database_path=/depot/itap/datasets/alphafold/db_20221014/mgnify/mgy_clusters_2018_12.fa
--uniclust30_database_path=/depot/itap/datasets/alphafold/db_20221014/uniclust30/uniclust30_2018_08/uniclust30_2018_08
--pdb_seqres_database_path=/depot/itap/datasets/alphafold/db_20221014/pdb_seqres/pdb_seqres.txt
--uniprot_database_path=/depot/itap/datasets/alphafold/db_20221014/uniprot/uniprot.fasta
--template_mmcif_dir=/depot/itap/datasets/alphafold/db_20221014/pdb_mmcif/mmcif_files
--obsolete_pdbs_path=/depot/itap/datasets/alphafold/db_20221014/pdb_mmcif/obsolete.dat
--hhblits_binary_path=/usr/bin/hhblits
--hhsearch_binary_path=/usr/bin/hhsearch
--jackhmmer_binary_path=/usr/bin/jackhmmer
--kalign_binary_path=/usr/bin/kalign

Note

Since Version v2.3.0, the AlphaFold-Multimer model parameters has been updated. The updated full database is stored in depot/itap/datasets/alphafold/db_20230311. For ACCESS Anvil, the database is stored in /anvil/datasets/alphafold/db_20230311. Users need to update the flagfile using the updated database:

run_alphafold.sh --flagfile=full_db_20230311.ff --fasta_paths=XX --output_dir=XX ...

Note

Since Version v2.3.0, uniclust30_database_path has been changed to uniref30_database_path.

full_db_20230311.ff (for alphafold v3)

Example contents of full_db_20230311.ff for monomer (For ACCESS Anvil, please change depot/itap to anvil):

--db_preset=full_dbs
--bfd_database_path=/depot/itap/datasets/alphafold/db_20230311/bfd/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt
--data_dir=/depot/itap/datasets/alphafold/db_20230311/
--uniref90_database_path=/depot/itap/datasets/alphafold/db_20230311/uniref90/uniref90.fasta
--mgnify_database_path=/depot/itap/datasets/alphafold/db_20230311/mgnify/mgy_clusters_2022_05.fa
--uniref30_database_path=/depot/itap/datasets/alphafold/db_20230311/uniref30/UniRef30_2021_03
--pdb70_database_path=/depot/itap/datasets/alphafold/db_20230311/pdb70/pdb70
--template_mmcif_dir=/depot/itap/datasets/alphafold/db_20230311/pdb_mmcif/mmcif_files
--obsolete_pdbs_path=/depot/itap/datasets/alphafold/db_20230311/pdb_mmcif/obsolete.dat
--hhblits_binary_path=/usr/bin/hhblits
--hhsearch_binary_path=/usr/bin/hhsearch
--jackhmmer_binary_path=/usr/bin/jackhmmer
--kalign_binary_path=/usr/bin/kalign

Example contents of full_db_20230311.ff for multimer (For ACCESS Anvil, please change depot/itap to anvil):

--db_preset=full_dbs
--bfd_database_path=/depot/itap/datasets/alphafold/db_20230311/bfd/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt
--data_dir=/depot/itap/datasets/alphafold/db_20230311/
--uniref90_database_path=/depot/itap/datasets/alphafold/db_20230311/uniref90/uniref90.fasta
--mgnify_database_path=/depot/itap/datasets/alphafold/db_20230311/mgnify/mgy_clusters_2022_05.fa
--uniref30_database_path=/depot/itap/datasets/alphafold/db_20230311/uniref30/UniRef30_2021_03
--pdb_seqres_database_path=/depot/itap/datasets/alphafold/db_20230311/pdb_seqres/pdb_seqres.txt
--uniprot_database_path=/depot/itap/datasets/alphafold/db_20230311/uniprot/uniprot.fasta
--template_mmcif_dir=/depot/itap/datasets/alphafold/db_20230311/pdb_mmcif/mmcif_files
--obsolete_pdbs_path=/depot/itap/datasets/alphafold/db_20230311/pdb_mmcif/obsolete.dat
--hhblits_binary_path=/usr/bin/hhblits
--hhsearch_binary_path=/usr/bin/hhsearch
--jackhmmer_binary_path=/usr/bin/jackhmmer
--kalign_binary_path=/usr/bin/kalign

Example job using CPU

Warning

Using #!/bin/sh -l as shebang in the slurm job script will cause the failure of some biocontainer modules. Please use #!/bin/bash instead.

Note

Notice that since version 2.2.0, the parameter --use_gpu_relax=False is required.

To run alphafold using CPU:

#!/bin/bash
#SBATCH -A myallocation     # Allocation name
#SBATCH -t 20:00:00
#SBATCH -N 1
#SBATCH -n 24
#SBATCH --job-name=alphafold
#SBATCH --mail-type=FAIL,BEGIN,END
#SBATCH --error=%x-%J-%u.err
#SBATCH --output=%x-%J-%u.out

module --force purge
ml biocontainers alphafold/2.3.1

run_alphafold.sh --flagfile=full_db_20230311.ff  \
    --fasta_paths=sample.fasta --max_template_date=2022-02-01 \
    --output_dir=af2_full_out --model_preset=monomer \
    --use_gpu_relax=False

Example job using GPU

Warning

Using #!/bin/sh -l as shebang in the slurm job script will cause the failure of some biocontainer modules. Please use #!/bin/bash instead.

Note

Notice that since version 2.2.0, the parameter --use_gpu_relax=True is required.

To run alphafold using GPU:

#!/bin/bash
#SBATCH -A myallocation     # Allocation name
#SBATCH -t 20:00:00
#SBATCH -N 1
#SBATCH -n 11
#SBATCH --gres=gpu:1
#SBATCH --job-name=alphafold
#SBATCH --mail-type=FAIL,BEGIN,END
#SBATCH --error=%x-%J-%u.err
#SBATCH --output=%x-%J-%u.out

module --force purge
ml biocontainers alphafold/2.3.1

run_alphafold.sh --flagfile=full_db_20230311.ff \
    --fasta_paths=sample.fasta --max_template_date=2022-02-01 \
    --output_dir=af2_full_out --model_preset=monomer \
    --use_gpu_relax=True