AlphaFold: Difference between revisions

From Alliance Doc
Jump to navigation Jump to search
No edit summary
Line 109: Line 109:


<!--T:15-->
<!--T:15-->
Note that this step <b>cannot</b> be done from a compute node. It should be done on a data transfer node (DTN) on clusters that have them (see [[Transferring data]]). On clusters that have no DTN, use a login node instead. Since the download can take up to a full day, we suggest using a [[Prolonging_terminal_sessions#Terminal_multiplexers|terminal multiplexer]]. You may encounter a <code>Client_loop: send disconnect: Broken pipe</code> error message even if you are using a terminal multiplexer. See [[AlphaFold#Broken pipe error message|Troubleshooting]].
Note that this step <b>cannot</b> be done from a compute node. It should be done on a data transfer node (DTN) on clusters that have them (see [[Transferring data]]). On clusters that have no DTN, use a login node instead. Since the download can take up to a full day, we suggest using a [[Prolonging_terminal_sessions#Terminal_multiplexers|terminal multiplexer]]. You may encounter a <code>Client_loop: send disconnect: Broken pipe</code> error message. See [[AlphaFold#Broken pipe error message|Troubleshooting]].


</tab>
</tab>

Revision as of 17:46, 17 November 2023

Other languages:

AlphaFold is a machine-learning model for the prediction of protein folding.

This page discusses how to use AlphaFold v2.0, the version that was entered in CASP14 and published in Nature.

Source code and documentation for AlphaFold can be found at their GitHub page. Any publication that discloses findings arising from using this source code or the model parameters should cite the AlphaFold paper.

Available versions

AlphaFold is available on our clusters as prebuilt Python packages (wheels). You can list available versions with avail_wheels.

Question.png 
[name@server ~]$ avail_wheels alphafold --all-versions
name       version    python    arch
---------  ---------  --------  -------
alphafold  2.3.1      py3       generic
alphafold  2.3.0      py3       generic
alphafold  2.2.4      py3       generic
alphafold  2.2.3      py3       generic
alphafold  2.2.2      py3       generic
alphafold  2.2.1      py3       generic
alphafold  2.1.1      py3       generic
alphafold  2.0.0      py3       generic

Installing AlphaFold in a Python virtual environment

1. Load AlphaFold dependencies.

Question.png 
[name@server ~]$ module load gcc/9.3.0 openmpi/4.0.3 cuda/11.4 cudnn/8.2.0 kalign/2.03 hmmer/3.2.1 openmm-alphafold/7.5.1 hh-suite/3.3.0 python/3.8

As of July 2022, only Python 3.7 and 3.8 are supported.


2. Create and activate a Python virtual environment. 

[name@server ~]$ virtualenv --no-download ~/alphafold_env
[name@server ~]$ source ~/alphafold_env/bin/activate


3. Install a specific version of AlphaFold and its Python dependencies. 

(alphafold_env) [name@server ~] pip install --no-index --upgrade pip
(alphafold_env) [name@server ~] pip install --no-index alphafold==X.Y.Z

where X.Y.Z is the exact desired version, for instance 2.2.4. You can omit to specify the version in order to install the latest one available from the wheelhouse.

4. Validate it.

Question.png 
(alphafold_env) [name@server ~] run_alphafold.py --help

5. Freeze the environment and requirements set.

Question.png 
(alphafold_env) [name@server ~] pip freeze > ~/alphafold-requirements.txt

Databases

Note that AlphaFold requires a set of databases.

A copy of the databases is available in /cvmfs/bio.data.computecanada.ca/content/databases/Core/alphafold2_dbs/2023_07/.

Question.png 
(alphafold_env) [name@server ~] export DOWNLOAD_DIR=/cvmfs/bio.data.computecanada.ca/content/databases/Core/alphafold2_dbs/2023_07/

Please note that the alphafold databases on cvmfs will undergo yearly updates as required. At present, the existing database is from July 2023.

The user can also choose to download the databases locally into their $SCRATCH directory.

Important: The databases must live in the $SCRATCH.

1. From a DTN or login node, create the data folder. 

(alphafold_env) [name@server ~] export DOWNLOAD_DIR=$SCRATCH/alphafold/data
(alphafold_env) [name@server ~] mkdir -p $DOWNLOAD_DIR


2. With your modules loaded and virtual environment activated, you can download the data.

Question.png 
(alphafold_env) [name@server ~] download_all_data.sh $DOWNLOAD_DIR

Note that this step cannot be done from a compute node. It should be done on a data transfer node (DTN) on clusters that have them (see Transferring data). On clusters that have no DTN, use a login node instead. Since the download can take up to a full day, we suggest using a terminal multiplexer. You may encounter a Client_loop: send disconnect: Broken pipe error message. See Troubleshooting.

1. Set DOWNLOAD_DIR.

Question.png 
(alphafold_env) [name@server ~] export DOWNLOAD_DIR=/datashare/alphafold


Afterwards, the structure of your data should be similar to

Question.png 
(alphafold_env) [name@server ~] tree -d $DOWNLOAD_DIR
$DOWNLOAD_DIR/                             # ~ 2.6 TB (total)
    bfd/                                   # ~ 1.8 TB
        # 6 files
    mgnify/                                # ~ 120 GB
        mgy_clusters.fa
    params/                                # ~ 5.3 GB
        # LICENSE
        # 15 models
        # 16 files (total)
    pdb70/                                 # ~ 56 GB
        # 9 files
    pdb_mmcif/                             # ~ 246 GB
        mmcif_files/
            # 202,764 files
        obsolete.dat
    pdb_seqres/                            # ~ 237 MB
        pdb_seqres.txt
    uniprot/                               # ~ 111 GB
        uniprot.fasta
    uniref30/                              # ~ 206 GB
        # 7 files
    uniref90/                              # ~ 73 GB
        uniref90.fasta
Question.png 
(alphafold_env) [name@server ~] tree -d $DOWNLOAD_DIR
$DOWNLOAD_DIR/                             # Total: ~ 2.2 TB (download: 428 GB)
    bfd/                                   # ~ 1.8 TB (download: 271.6 GB)
        # 6 files.
    mgnify/                                # ~ 64 GB (download: 32.9 GB)
        mgy_clusters.fa
    params/                                # ~ 3.5 GB (download: 3.5 GB)
        # 5 CASP14 models,
        # 5 pTM models,
        # LICENSE,
        # = 11 files.
    pdb70/                                 # ~ 56 GB (download: 19.5 GB)
        # 9 files.
    pdb_mmcif/                             # ~ 206 GB (download: 46 GB)
        mmcif_files/
            # About 180,000 .cif files.
        obsolete.dat
    uniclust30/                            # ~ 87 GB (download: 24.9 GB)
        uniclust30_2018_08/
            # 13 files.
    uniref90/                              # ~ 59 GB (download: 29.7 GB)
        uniref90.fasta

Running AlphaFold

Performance

You can request at most 8 CPU cores when running AlphaFold because it is hardcoded to not use more and does not benefit from using more.



Edit one of following submission scripts according to your needs.

File : alphafold-2.3-cpu.sh

#!/bin/bash

#SBATCH --job-name=alphafold_run
#SBATCH --account=def-someprof    # adjust this to match the accounting group you are using to submit jobs
#SBATCH --time=08:00:00           # adjust this to match the walltime of your job
#SBATCH --cpus-per-task=8         # a MAXIMUM of 8 core, Alpafold has no benefit to use more
#SBATCH --mem=20G                 # adjust this according to the memory you need

# Load modules dependencies
module load gcc/9.3.0 openmpi/4.0.3 cuda/11.4 cudnn/8.2.0 kalign/2.03 hmmer/3.2.1 openmm-alphafold/7.5.1 hh-suite/3.3.0 python/3.8

DOWNLOAD_DIR=$SCRATCH/alphafold/data   # set the appropriate path to your downloaded data
INPUT_DIR=$SCRATCH/alphafold/input     # set the appropriate path to your input data
OUTPUT_DIR=${SCRATCH}/alphafold/output # set the appropriate path to your output data

# Generate your virtual environment in $SLURM_TMPDIR
virtualenv --no-download ${SLURM_TMPDIR}/env
source ${SLURM_TMPDIR}/env/bin/activate

# Install alphafold and its dependencies
pip install --no-index --upgrade pip
pip install --no-index --requirement ~/alphafold-requirements.txt

# Edit with the proper arguments, run your commands.
# run_alphafold.py --help
run_alphafold.py \
   --fasta_paths=${INPUT_DIR}/YourSequence.fasta,${INPUT_DIR}/AnotherSequence.fasta \
   --output_dir=${OUTPUT_DIR} \
   --data_dir=${DOWNLOAD_DIR} \
   --db_preset=full_dbs \
   --model_preset=multimer \
   --bfd_database_path=${DOWNLOAD_DIR}/bfd/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt \
   --mgnify_database_path=${DOWNLOAD_DIR}/mgnify/mgy_clusters_2022_05.fa \
   --pdb70_database_path=${DOWNLOAD_DIR}/pdb70/pdb70 \
   --template_mmcif_dir=${DOWNLOAD_DIR}/pdb_mmcif/mmcif_files \
   --obsolete_pdbs_path=${DOWNLOAD_DIR}/pdb_mmcif/obsolete.dat \
   --pdb_seqres_database_path=${DOWNLOAD_DIR}/pdb_seqres/pdb_seqres.txt \
   --uniprot_database_path=${DOWNLOAD_DIR}/uniprot/uniprot.fasta \
   --uniref30_database_path=${DOWNLOAD_DIR}/uniref30/UniRef30_2021_03 \
   --uniref90_database_path=${DOWNLOAD_DIR}/uniref90/uniref90.fasta \
   --hhblits_binary_path=${EBROOTHHMINSUITE}/bin/hhblits \
   --hhsearch_binary_path=${EBROOTHHMINSUITE}/bin/hhsearch \
   --jackhmmer_binary_path=${EBROOTHMMER}/bin/jackhmmer \
   --kalign_binary_path=${EBROOTKALIGN}/bin/kalign \
   --max_template_date=2021-12-31 \
   --use_gpu_relax=False


File : alphafold-2.3-gpu.sh

#!/bin/bash

#SBATCH --job-name=alphafold_run
#SBATCH --account=def-someprof    # adjust this to match the accounting group you are using to submit jobs
#SBATCH --time=08:00:00           # adjust this to match the walltime of your job
#SBATCH --cpus-per-task=8         # a MAXIMUM of 8 core, Alpafold has no benefit to use more
#SBATCH --gres=gpu:1              # a GPU helps to accelerate the inference part only
#SBATCH --mem=20G                 # adjust this according to the memory you need

# Load modules dependencies
module load gcc/9.3.0 openmpi/4.0.3 cuda/11.4 cudnn/8.2.0 kalign/2.03 hmmer/3.2.1 openmm-alphafold/7.5.1 hh-suite/3.3.0 python/3.8

DOWNLOAD_DIR=$SCRATCH/alphafold/data   # set the appropriate path to your downloaded data
INPUT_DIR=$SCRATCH/alphafold/input     # set the appropriate path to your input data
OUTPUT_DIR=${SCRATCH}/alphafold/output # set the appropriate path to your output data

# Generate your virtual environment in $SLURM_TMPDIR
virtualenv --no-download ${SLURM_TMPDIR}/env
source ${SLURM_TMPDIR}/env/bin/activate

# Install alphafold and its dependencies
pip install --no-index --upgrade pip
pip install --no-index --requirement ~/alphafold-requirements.txt

# Edit with the proper arguments, run your commands.
# run_alphafold.py --help
run_alphafold.py \
   --fasta_paths=${INPUT_DIR}/YourSequence.fasta,${INPUT_DIR}/AnotherSequence.fasta \
   --output_dir=${OUTPUT_DIR} \
   --data_dir=${DOWNLOAD_DIR} \
   --db_preset=full_dbs \
   --model_preset=multimer \
   --bfd_database_path=${DOWNLOAD_DIR}/bfd/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt \
   --mgnify_database_path=${DOWNLOAD_DIR}/mgnify/mgy_clusters_2022_05.fa \
   --pdb70_database_path=${DOWNLOAD_DIR}/pdb70/pdb70 \
   --template_mmcif_dir=${DOWNLOAD_DIR}/pdb_mmcif/mmcif_files \
   --obsolete_pdbs_path=${DOWNLOAD_DIR}/pdb_mmcif/obsolete.dat \
   --pdb_seqres_database_path=${DOWNLOAD_DIR}/pdb_seqres/pdb_seqres.txt \
   --uniprot_database_path=${DOWNLOAD_DIR}/uniprot/uniprot.fasta \
   --uniref30_database_path=${DOWNLOAD_DIR}/uniref30/UniRef30_2021_03 \
   --uniref90_database_path=${DOWNLOAD_DIR}/uniref90/uniref90.fasta \
   --hhblits_binary_path=${EBROOTHHMINSUITE}/bin/hhblits \
   --hhsearch_binary_path=${EBROOTHHMINSUITE}/bin/hhsearch \
   --jackhmmer_binary_path=${EBROOTHMMER}/bin/jackhmmer \
   --kalign_binary_path=${EBROOTKALIGN}/bin/kalign \
   --max_template_date=2021-12-31 \
   --use_gpu_relax=True


File : alphafold-cpu.sh

#!/bin/bash

#SBATCH --job-name=alphafold_run
#SBATCH --account=def-someprof    # adjust this to match the accounting group you are using to submit jobs
#SBATCH --time=08:00:00           # adjust this to match the walltime of your job
#SBATCH --cpus-per-task=8         # a MAXIMUM of 8 core, Alpafold has no benefit to use more
#SBATCH --mem=20G                 # adjust this according to the memory you need

# Load modules dependencies
module load gcc/9.3.0 openmpi/4.0.3 cuda/11.4 cudnn/8.2.0 kalign/2.03 hmmer/3.2.1 openmm-alphafold/7.5.1 hh-suite/3.3.0 python/3.8

DOWNLOAD_DIR=$SCRATCH/alphafold/data   # set the appropriate path to your downloaded data
INPUT_DIR=$SCRATCH/alphafold/input     # set the appropriate path to your input data
OUTPUT_DIR=${SCRATCH}/alphafold/output # set the appropriate path to your output data

# Generate your virtual environment in $SLURM_TMPDIR
virtualenv --no-download ${SLURM_TMPDIR}/env
source ${SLURM_TMPDIR}/env/bin/activate

# Install alphafold and its dependencies
pip install --no-index --upgrade pip
pip install --no-index --requirement ~/alphafold-requirements.txt

# Edit with the proper arguments, run your commands.
# Note that the `--uniclust30_database_path` option below was renamed to
# `--uniref30_database_path` in 2.3.
# run_alphafold.py --help
run_alphafold.py \
   --fasta_paths=${INPUT_DIR}/YourSequence.fasta,${INPUT_DIR}/AnotherSequence.fasta \
   --output_dir=${OUTPUT_DIR} \
   --data_dir=${DOWNLOAD_DIR} \
   --model_preset=monomer_casp14 \
   --bfd_database_path=${DOWNLOAD_DIR}/bfd/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt \
   --mgnify_database_path=${DOWNLOAD_DIR}/mgnify/mgy_clusters_2018_12.fa \
   --pdb70_database_path=${DOWNLOAD_DIR}/pdb70/pdb70 \
   --template_mmcif_dir=${DOWNLOAD_DIR}/pdb_mmcif/mmcif_files \
   --obsolete_pdbs_path=${DOWNLOAD_DIR}/pdb_mmcif/obsolete.dat \
   --uniclust30_database_path=${DOWNLOAD_DIR}/uniclust30/uniclust30_2018_08/uniclust30_2018_08  \
   --uniref90_database_path=${DOWNLOAD_DIR}/uniref90/uniref90.fasta  \
   --hhblits_binary_path=${EBROOTHHMINSUITE}/bin/hhblits \
   --hhsearch_binary_path=${EBROOTHHMINSUITE}/bin/hhsearch \
   --jackhmmer_binary_path=${EBROOTHMMER}/bin/jackhmmer \
   --kalign_binary_path=${EBROOTKALIGN}/bin/kalign \
   --max_template_date=2020-05-14 \
   --use_gpu_relax=False


File : alphafold-gpu.sh

#!/bin/bash

#SBATCH --job-name=alphafold_run
#SBATCH --account=def-someprof    # adjust this to match the accounting group you are using to submit jobs
#SBATCH --time=08:00:00           # adjust this to match the walltime of your job
#SBATCH --gres=gpu:1              # a GPU helps to accelerate the inference part only
#SBATCH --cpus-per-task=8         # a MAXIMUM of 8 core, Alpafold has no benefit to use more
#SBATCH --mem=20G                 # adjust this according to the memory you need

# Load modules dependencies
module load gcc/9.3.0 openmpi/4.0.3 cuda/11.4 cudnn/8.2.0 kalign/2.03 hmmer/3.2.1 openmm-alphafold/7.5.1 hh-suite/3.3.0 python/3.8

DOWNLOAD_DIR=$SCRATCH/alphafold/data   # set the appropriate path to your downloaded data
INPUT_DIR=$SCRATCH/alphafold/input     # set the appropriate path to your input data
OUTPUT_DIR=${SCRATCH}/alphafold/output # set the appropriate path to your output data

# Generate your virtual environment in $SLURM_TMPDIR
virtualenv --no-download ${SLURM_TMPDIR}/env
source ${SLURM_TMPDIR}/env/bin/activate

# Install alphafold and its dependencies
pip install --no-index --upgrade pip
pip install --no-index --requirement ~/alphafold-requirements.txt

# Edit with the proper arguments, run your commands.
# Note that the `--uniclust30_database_path` option below was renamed to
# `--uniref30_database_path` in 2.3.
# run_alphafold.py --help
run_alphafold.py \
   --fasta_paths=${INPUT_DIR}/YourSequence.fasta,${INPUT_DIR}/AnotherSequence.fasta \
   --output_dir=${OUTPUT_DIR} \
   --data_dir=${DOWNLOAD_DIR} \
   --model_preset=monomer_casp14 \
   --bfd_database_path=${DOWNLOAD_DIR}/bfd/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt \
   --mgnify_database_path=${DOWNLOAD_DIR}/mgnify/mgy_clusters_2018_12.fa \
   --pdb70_database_path=${DOWNLOAD_DIR}/pdb70/pdb70 \
   --template_mmcif_dir=${DOWNLOAD_DIR}/pdb_mmcif/mmcif_files \
   --obsolete_pdbs_path=${DOWNLOAD_DIR}/pdb_mmcif/obsolete.dat \
   --uniclust30_database_path=${DOWNLOAD_DIR}/uniclust30/uniclust30_2018_08/uniclust30_2018_08  \
   --uniref90_database_path=${DOWNLOAD_DIR}/uniref90/uniref90.fasta  \
   --hhblits_binary_path=${EBROOTHHMINSUITE}/bin/hhblits \
   --hhsearch_binary_path=${EBROOTHHMINSUITE}/bin/hhsearch \
   --jackhmmer_binary_path=${EBROOTHMMER}/bin/jackhmmer \
   --kalign_binary_path=${EBROOTKALIGN}/bin/kalign \
   --max_template_date=2020-05-14 \
   --use_gpu_relax=True


Then, submit the job to the scheduler.

Question.png 
(alphafold_env) [name@server ~] sbatch --job-name alphafold-X alphafold-gpu.sh

Troubleshooting

Broken pipe error message

When downloading the database, you may encounter a Client_loop: send disconnect: Broken pipe error message even if you are using a terminal multiplexer. It is hard to find the exact cause that is responsible for this error message. It could be as simple as an unusually high number of users working on the login node giving less space for you to upload data.

  • One solution would be to use the database that is already present on the clusters : /cvmfs/bio.data.computecanada.ca/content/databases/Core/alphafold2_dbs/2023_07/.
  • Another option would be to download the full database in sections. To have access to the different download scripts, after loading the module and activated your virtual environment, you simply enter download_ in your terminal and tap twice on the tab keyboard key to visualize all the scripts that are available. You can manually download sections of the database by using the available script as for instance : download_pdb.sh.
Retrieved from "https://docs.alliancecan.ca/mediawiki/index.php?title=AlphaFold&oldid=146637"