AlphaFold: Difference between revisions

From Alliance Doc
Jump to navigation Jump to search
m (Specify correct modules versions and names)
m (move Category tag outside of translated area of the page)
 
(147 intermediate revisions by 13 users not shown)
Line 1: Line 1:
{{draft}}
<languages />
This package provides an implementation of the inference pipeline of AlphaFold v2.0. This is a completely new model that was entered in CASP14 and published in Nature. For simplicity, we refer to this model as AlphaFold throughout the rest of this document.
[[Category:Software]]


Any publication that discloses findings arising from using this source code or the model parameters should [https://github.com/deepmind/alphafold#citing-this-work cite] the [https://doi.org/10.1038/s41586-021-03819-2 AlphaFold paper].
<translate>
<!--T:1-->
[https://deepmind.com/blog/article/alphafold-a-solution-to-a-50-year-old-grand-challenge-in-biology AlphaFold]
is a machine learning model for the prediction of protein folding.  


The source code of this package can be found in their [https://github.com/deepmind/alphafold GitHub page] along with some documentation.
<!--T:2-->
This page discusses how to use AlphaFold v2.0, the version that was entered in CASP14 and published in Nature.


== Usage in Compute Canada systems ==
<!--T:3-->
As you might have seen from their documentation, they explain the usage via Docker. In Compute Canada we do not provide Docker as container, but singularity (see our documentation at https://docs.computecanada.ca/wiki/Singularity). However, we have created a wheel to use AlphaFold in a python environment.
Source code and documentation for AlphaFold can be found at their [https://github.com/deepmind/alphafold GitHub page].
Any publication that discloses findings arising from use of this source code or the model parameters should [https://github.com/deepmind/alphafold#citing-this-work cite] the [https://doi.org/10.1038/s41586-021-03819-2 AlphaFold paper].


=== AlphaFold in Python environment ===
== Available versions == <!--T:5-->
1. AlphaFold has a number of non-python dependencies that need to be loaded ahead of time. For example, cuda, kalign, hmmer, and openmm. Luckily all these dependencies are available though our stack:
AlphaFold is available on our clusters as prebuilt Python packages (wheels). You can list available versions with <code>avail_wheels</code>.
{{Command
|avail_wheels alphafold --all-versions
|result=
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
}}


<pre>
== Installing AlphaFold in a Python virtual environment == <!--T:6-->
[name@cluster ~]$ module load gcc/9 openmpi cuda/11.1 cudnn/8.2.0 kalign/2.03 hmmer/3.2.1 openmm-alphafold/7.5.1 hh-suite/3.3.0 python/3.8
</pre>


2. Then you can proceed to create the python virtual environment and activate it by:
<!--T:7-->
1. Load AlphaFold dependencies.
{{Command|module load StdEnv/2020 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.


<pre>
[name@cluster ~]$ virtualenv --no-download ~/my_env && source ~/my_env/bin/activate
</pre>


3. Now you can install AlphaFold and its dependencies by:
<!--T:8-->
2. Create and activate a Python virtual environment.
{{Commands
|virtualenv --no-download ~/alphafold_env
|source ~/alphafold_env/bin/activate
}}


<pre>
<!--T:9-->
(my_env)[name@cluster ~]$ pip install --no-index six==1.15 numpy==1.19.2 scipy==1.4.1 pdbfixer alphafold
3. Install a specific version of AlphaFold and its Python dependencies.
</pre>
{{Commands
|prompt=(alphafold_env) [name@server ~]
|pip install --no-index --upgrade pip
|pip install --no-index alphafold{{=}}{{=}}X.Y.Z
}}
where <code>X.Y.Z</code> is the exact desired version, for instance <code>2.2.4</code>.
You can omit to specify the version in order to install the latest one available from the wheelhouse.


Now AlphaFold is ready to be used.
<!--T:10-->
4. Validate it.
{{Command
|prompt=(alphafold_env) [name@server ~]
|run_alphafold.py --help
}}


==== Creating the virtual environment in the jobscript ====
<!--T:45-->
As you probably have read in [Creating_and_using_a_virtual_environment](https://docs.computecanada.ca/wiki/Python#Creating_and_using_a_virtual_environment), you can also take advantage of the local installs on compute nodes:
5. Freeze the environment and requirements set.
{{Command
|prompt=(alphafold_env) [name@server ~]
|pip freeze > ~/alphafold-requirements.txt
}}
 
== Databases == <!--T:11-->
Note that AlphaFold requires a set of databases.
 
<!--T:65-->
The databases are available in
<code>/cvmfs/bio.data.computecanada.ca/content/databases/Core/alphafold2_dbs/</code>.


<!--T:63-->
AlphaFold databases on CVMFS undergo yearly updates. In January 2024, the database was updated and is accessible in folder <code>2024_01</code>.
{{Command
|prompt=(alphafold_env) [name@server ~]
|export DOWNLOAD_DIR{{=}}/cvmfs/bio.data.computecanada.ca/content/databases/Core/alphafold2_dbs/2024_01/
}}


{{File
<!--T:66-->
|name=my_alphafoldjob.sh
You can also choose to download the databases locally into your <code>$SCRATCH</code> directory.
|lang="bash"
 
|contents=
<!--T:12-->
#!/bin/bash
<b>Important:</b> The databases must live in the <code>$SCRATCH</code> directory.
#SBATCH --job-name=alphafold_run
 
#SBATCH --account=def-someprof # adjust this to match the accounting group you are using to submit jobs
<!--T:13-->
#SBATCH --time=0-03:00        # adjust this to match the walltime of your job
<tabs>
#SBATCH --nodes=1     
<tab name="General">
#SBATCH --ntasks=1
1. From a DTN or login node, create the data folder.
#SBATCH --gres=gpu:1          # You need to request one GPU to be able to run AlphaFold properly
{{Commands
#SBATCH --cpus-per-task=1     # adjust this if you are using parallel commands
|prompt=(alphafold_env) [name@server ~]
#SBATCH --mem=4000            # adjust this according to the memory requirement per node you need
|export DOWNLOAD_DIR{{=}}$SCRATCH/alphafold/data
#SBATCH --mail-user=you@youruniversity.ca # adjust this to match your email address
|mkdir -p $DOWNLOAD_DIR
#SBATCH --mail-type=ALL
}}


# Load your modules as before
<!--T:14-->
module load gcc/9 openmpi cuda/11.1 cudnn/8.2.0 kalign/2.03 hmmer/3.2.1 openmm-alphafold/7.5.1 hh-suite/3.3.0 python/3.8
2. With your modules loaded and virtual environment activated, you can download the data.
{{Command
|prompt=(alphafold_env) [name@server ~]
|download_all_data.sh $DOWNLOAD_DIR
}}


# Generate your virtual environment in $SLURM_TMPDIR
<!--T:15-->
virtualenv --no-download ${SLURM_TMPDIR}/my_env && source ${SLURM_TMPDIR}/my_env/bin/activate
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]] below.


# Install alphafold and dependencies
<!--T:67-->
pip install --no-index six==1.15 numpy==1.19.2 scipy==1.4.1 pdbfixer
</tab>


# Run your commands
<!--T:16-->
python run_alphafold.py --help
<tab name="Graham only">
1. Set <code>DOWNLOAD_DIR</code>.
{{Command
|prompt=(alphafold_env) [name@server ~]
|export DOWNLOAD_DIR{{=}}/datashare/alphafold
}}
}}


== Databases ==
<!--T:62-->
Note that AlphaFold requires a set of datasets/databases that need to be downloaded. Also notice that we prefer you avoid using `aria2c`. For your convenience we have included a modified version of the scripts within the AlphaFold wheel, so within the virtual environment you can do:
</tab>
</tabs>


<pre>
(my_env)[name@cluster ~]$ download_all_data.sh path/to/download
</pre>


Note that this might take a while and '''SHOULD NOT BE DONE IN THE COMPUTE NODES'''. Instead, you should use the [https://docs.computecanada.ca/wiki/Transferring_data data transfer nodes] or the login nodes. Since the download might take a while we recommend you do this in a [https://linuxize.com/post/how-to-use-linux-screen/ screen] or [https://docs.computecanada.ca/wiki/Tmux Tmux] sessions. If your path/to/download is stored in `$DOWNLOAD_DIR`, then the structure of your data should be:
<!--T:47-->
Afterwards, the structure of your data should be similar to
<tabs>
<tab name=2.3>
{{Command
|prompt=(alphafold_env) [name@server ~]
|tree -d $DOWNLOAD_DIR
|result=
$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
}}
</tab>


<pre>
<!--T:17-->
<tab name=2.2>
{{Command
|prompt=(alphafold_env) [name@server ~]
|tree -d $DOWNLOAD_DIR
|result=
$DOWNLOAD_DIR/                            # Total: ~ 2.2 TB (download: 428 GB)
$DOWNLOAD_DIR/                            # Total: ~ 2.2 TB (download: 428 GB)
     bfd/                                  # ~ 1.8 TB (download: 271.6 GB)
     bfd/                                  # ~ 1.8 TB (download: 271.6 GB)
Line 94: Line 188:
     uniref90/                              # ~ 59 GB (download: 29.7 GB)
     uniref90/                              # ~ 59 GB (download: 29.7 GB)
         uniref90.fasta
         uniref90.fasta
</pre>
}}
</tab>
</tabs>


This is important when passing the commands to AlphaFold.
== Running AlphaFold == <!--T:18-->
{{Warning
|title=Performance
|content=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.
}}


== Running AlphaFold ==
<!--T:19-->
Once you have everything setup, you can run a production run of AlphaFold by:
Edit one of following submission scripts according to your needs.
<tabs>
<tab name="2.3 on CPU">
{{File
|name=alphafold-2.3-cpu.sh
|lang="bash"
|contents=
#!/bin/bash
 
<!--T:48-->
#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, AlphaFold has no benefit to use more
#SBATCH --mem=20G                # adjust this according to the memory you need
 
<!--T:49-->
# Load modules dependencies.
module load StdEnv/2020 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
 
<!--T:50-->
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


<!--T:51-->
# Generate your virtual environment in $SLURM_TMPDIR.
virtualenv --no-download ${SLURM_TMPDIR}/env
source ${SLURM_TMPDIR}/env/bin/activate
<!--T:52-->
# Install AlphaFold and its dependencies.
pip install --no-index --upgrade pip
pip install --no-index --requirement ~/alphafold-requirements.txt
<!--T:53-->
# Edit with the proper arguments and run your commands.
# run_alphafold.py --help
run_alphafold.py \
  --fasta_paths=${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=2022-01-01 \
  --use_gpu_relax=False
}}
</tab>
<!--T:54-->
<tab name="2.3 on GPU">
{{File
{{File
|name=my_alphafoldjob.sh
|name=alphafold-2.3-gpu.sh
|lang="bash"
|lang="bash"
|contents=
|contents=
#!/bin/bash
#!/bin/bash
<!--T:55-->
#SBATCH --job-name=alphafold_run
#SBATCH --job-name=alphafold_run
#SBATCH --account=def-someprof # adjust this to match the accounting group you are using to submit jobs
#SBATCH --account=def-someprof   # adjust this to match the accounting group you are using to submit jobs
#SBATCH --time=0-03:00         # adjust this to match the walltime of your job
#SBATCH --time=08:00:00           # adjust this to match the walltime of your job
#SBATCH --nodes=1     
#SBATCH --cpus-per-task=8        # a MAXIMUM of 8 core, AlphaFold has no benefit to use more
#SBATCH --ntasks=1
#SBATCH --gres=gpu:1             # a GPU helps to accelerate the inference part only
#SBATCH --gres=gpu:1           # You need to request one GPU to be able to run AlphaFold properly
#SBATCH --mem=20G                # adjust this according to the memory you need
#SBATCH --cpus-per-task=8      # adjust this if you are using parallel commands
#SBATCH --mem=32G              # adjust this according to the memory requirement per node you need
#SBATCH --mail-user=you@youruniversity.ca # adjust this to match your email address
#SBATCH --mail-type=ALL


# Set the path to download dir
<!--T:56-->
DOWNLOAD_DIR=/projects/def-someprof/${USER}/data  # Set the appropriate path to your downloaded data
# Load modules dependencies.
DATA_DIR=/projects/def-someprof/${USER}/input    # Set the appropriate path to your supporting data
module load StdEnv/2020 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


# Load your modules as before
<!--T:57-->
module load gcc openmpi cuda/11.1 cudacore/.11.1.1 cudnn/8.2.0 kalign hmmer hh-suite openmm python/3.7
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
<!--T:58-->
virtualenv --no-download ${SLURM_TMPDIR}/my_env && source ${SLURM_TMPDIR}/my_env/bin/activate
# Generate your virtual environment in $SLURM_TMPDIR.
virtualenv --no-download ${SLURM_TMPDIR}/env
source ${SLURM_TMPDIR}/env/bin/activate


# Install alphafold and dependencies
<!--T:59-->
pip install --no-index six==1.15 numpy==1.19.2 scipy==1.4.1 pdbfixer alphafold
# Install AlphaFold and its dependencies.
pip install --no-index --upgrade pip
pip install --no-index --requirement ~/alphafold-requirements.txt


# Run your commands
<!--T:60-->
python run_alphafold.py --bfd_database_path ${DOWNLOAD_DIR}/bfd --data_dir ${DATA_DIR} \
# Edit with the proper arguments and run your commands.
   --fasta_paths ${DATA_DIR}/fasta1.fasta,${DATA_DIR}/fasta2.fasta,${DATA_DIR}/fasta3.fasta \
# run_alphafold.py --help
   --hhblits_binary_path ${EBROOTHHMINSUITE}/bin/hhblits \
run_alphafold.py \
   --hhsearch_binary_path ${EBROOTHHMINSUITE}/bin//hhsearch')
  --fasta_paths=${INPUT_DIR}/YourSequence.fasta,${INPUT_DIR}/AnotherSequence.fasta \
   --jackhmmer_binary_path ${EBROOTHMMER}/bin/jackhmmer \
  --output_dir=${OUTPUT_DIR} \
   --kalign_binary_path ${EBROOTKALIGN}/bin/kalign
  --data_dir=${DOWNLOAD_DIR} \
   --mgnify_database_path ${DOWNLOAD_DIR}/mgnify --model_names model1,model2 \ # use the actual models you want
   --db_preset=full_dbs \
   --output_dir ~/scratch/alphafold_output --pdb70_database_path ${DOWNLOAD_DIR}/pdb70 \
  --model_preset=multimer \
   --template_mmcif_dir ${DATA_DIR}/Templates --uniclust30_database_path ${DOWNLOAD_DIR}/uniclust30
  --bfd_database_path=${DOWNLOAD_DIR}/bfd/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt \
   --uniref90_database_path ${DOWNLOAD_DIR}/uniref90
  --mgnify_database_path=${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=${EBROOTKALIGN}/bin/kalign \
  --max_template_date=2022-01-01 \
  --use_gpu_relax=True
}}
}}
</tab>
<!--T:61-->
<tab name="2.2 on CPU">
{{File
|name=alphafold-cpu.sh
|lang="bash"
|contents=
#!/bin/bash
<!--T:20-->
#SBATCH --job-name=alphafold_run
#SBATCH --account=def-someprof    # adjust this to match the accounting group you are using to submit jobs
#SBATCH --time=08:00:00          # adjust this to match the walltime of your job
#SBATCH --cpus-per-task=8        # a MAXIMUM of 8 core, AlphaFold has no benefit to use more
#SBATCH --mem=20G                # adjust this according to the memory you need
<!--T:21-->
# Load modules dependencies.
module load StdEnv/2020 gcc/9.3.0 openmpi/4.0.3 cuda/11.4 cudnn/8.2.0 kalign/2.03 hmmer/3.2.1 openmm-alphafold/7.5.1 hh-suite/3.3.0 python/3.8
<!--T:22-->
DOWNLOAD_DIR=$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
<!--T:23-->
# Generate your virtual environment in $SLURM_TMPDIR.
virtualenv --no-download ${SLURM_TMPDIR}/env
source ${SLURM_TMPDIR}/env/bin/activate
<!--T:24-->
# Install AlphaFold and its dependencies.
pip install --no-index --upgrade pip
pip install --no-index --requirement ~/alphafold-requirements.txt
<!--T:25-->
# Edit with the proper arguments and run your commands.
# Note that the `--uniclust30_database_path` option below was renamed to
# `--uniref30_database_path` in 2.3.
# run_alphafold.py --help
run_alphafold.py \
  --fasta_paths=${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
}}
</tab>
<!--T:26-->
<tab name="2.2 on GPU">
{{File
|name=alphafold-gpu.sh
|lang="bash"
|contents=
#!/bin/bash
<!--T:27-->
#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, AlphaFold has no benefit to use more
#SBATCH --mem=20G                # adjust this according to the memory you need
<!--T:28-->
# Load modules dependencies.
module load StdEnv/2020 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
<!--T:29-->
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
<!--T:30-->
# Generate your virtual environment in $SLURM_TMPDIR.
virtualenv --no-download ${SLURM_TMPDIR}/env
source ${SLURM_TMPDIR}/env/bin/activate
<!--T:31-->
# Install AlphaFold  and its dependencies.
pip install --no-index --upgrade pip
pip install --no-index --requirement ~/alphafold-requirements.txt
<!--T:32-->
# Edit with the proper arguments and run your commands.
# Note that the `--uniclust30_database_path` option below was renamed to
# `--uniref30_database_path` in 2.3.
# run_alphafold.py --help
run_alphafold.py \
  --fasta_paths=${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
}}
</tab>
</tabs>
<!--T:33-->
Then, submit the job to the scheduler.
{{Command
|prompt=(alphafold_env) [name@server ~]
|sbatch --job-name alphafold-X alphafold-gpu.sh
}}
== Troubleshooting == <!--T:68-->
=== Broken pipe error message ===
When downloading the database, you may encounter a <code>Client_loop: send disconnect: Broken pipe</code> error message. It is hard to find the exact cause for this error message. It could be as simple as an unusually high number of users working on the login node, leaving less space for you to upload data.


== Using singularity ==
<!--T:69-->
In case you want to try the conternarized version (NOT our preferred option), you can [https://docs.computecanada.ca/wiki/Singularity/en#Creating_images_on_Compute_Canada_clusters build a singularity container]:
*One solution is to use a [[Prolonging_terminal_sessions#Terminal_multiplexers|terminal multiplexer]]. Note that you could still encounter this error message but less are the chances.


<pre>
<!--T:70-->
[name@cluster ~]$ module load singularity
*A second solution is to use the database that is already present on the cluster. <code>/cvmfs/bio.data.computecanada.ca/content/databases/Core/alphafold2_dbs/2023_07/</code>.
[name@cluster ~]$ singularity build alphafold.sif docker://uvarc/alphafold:2.0.0
</pre>


Before trying to build it or run it, check our [https://docs.computecanada.ca/wiki/Singularity singularity documentation] as there are particularities of each system that need to be taken into account.
<!--T:71-->
*Another option is 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 <code>download_</code> in your terminal and tap twice on the <code>tab</code> 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 <code>download_pdb.sh</code>.  


=== Running AlphaFold within Singularity ===
</translate>

Latest revision as of 12:47, 1 May 2024

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 use of this source code or the model parameters should cite the AlphaFold paper.

Available versions[edit]

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[edit]

1. Load AlphaFold dependencies.

Question.png 
[name@server ~]$ module load StdEnv/2020 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[edit]

Note that AlphaFold requires a set of databases.

The databases are available in /cvmfs/bio.data.computecanada.ca/content/databases/Core/alphafold2_dbs/.

AlphaFold databases on CVMFS undergo yearly updates. In January 2024, the database was updated and is accessible in folder 2024_01.

Question.png 
(alphafold_env) [name@server ~] export DOWNLOAD_DIR=/cvmfs/bio.data.computecanada.ca/content/databases/Core/alphafold2_dbs/2024_01/

You can also choose to download the databases locally into your $SCRATCH directory.

Important: The databases must live in the $SCRATCH directory.

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 below.

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[edit]

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, AlphaFold has no benefit to use more
#SBATCH --mem=20G                 # adjust this according to the memory you need

# Load modules dependencies.
module load StdEnv/2020 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 and 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=2022-01-01 \
   --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, AlphaFold 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 StdEnv/2020 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 and 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=2022-01-01 \
   --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, AlphaFold has no benefit to use more
#SBATCH --mem=20G                 # adjust this according to the memory you need

# Load modules dependencies.
module load StdEnv/2020 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 and 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, AlphaFold has no benefit to use more
#SBATCH --mem=20G                 # adjust this according to the memory you need

# Load modules dependencies.
module load StdEnv/2020 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 and 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[edit]

Broken pipe error message[edit]

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

  • One solution is to use a terminal multiplexer. Note that you could still encounter this error message but less are the chances.
  • A second solution is to use the database that is already present on the cluster. /cvmfs/bio.data.computecanada.ca/content/databases/Core/alphafold2_dbs/2023_07/.
  • Another option is 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.

CernVM File System

Retrieved from "https://docs.alliancecan.ca/mediawiki/index.php?title=AlphaFold&oldid=152755"