What to expect after running installation and do you have any test data to check if it's working? #553

Activity

1. DimaMolod commented on Nov 18, 2025

   

   DimaMolod

   on Nov 18, 2025

   Collaborator

   More actions

   Hi @MauriAndresMU1313!
   after snakedeploy, you will get structure like this:

   |-- config
   |   |-- config.yaml
   |   |-- profiles
   |   |   |-- desktop
   |   |       `-- config.yaml
   |   |   |-- slurm
   |   |       `-- config.yaml
   |   |-- sample_sheet.csv
   |   `-- split_fasta.py
   `-- workflow
       `-- Snakefile

   including the configurations file AlphaPulldownSnakemake/config/config.yaml where you need to specify whether you want to use alphafold2 or alphafold3 (for features and predictions), add paths to your AF databases on HPC, alphafold weights, output directory, csv file with complexes to be modeled, etc.
   The version of config.yaml you get is configured for EMBL cluster, which will not work out-of-the-box for other clusters.
   To check that you configured your run correctly, you can run dry run with
   snakemake -n
   after running AlphaFold2 (for AF3 layout is slightly different), you'll see output dir like that:

   |-- output_directory
   |   |-- data
   |   |   |-- P01258.fasta
   |   |   `-- P01579.fasta
   |   |-- features
   |   |   |-- P01258_feature_metadata_2025-11-17.json.xz
   |   |   |-- P01258.pkl.xz
   |   |   |-- P01579_feature_metadata_2025-11-17.json.xz
   |   |   `-- P01579.pkl.xz
   |   `-- predictions
   |       |-- P01258+P01579
   |       |-- ...
   |-- slurm-logs
   |   |-- ...
   |   |-- reports
   |       `-- all_interfaces.csv

   Please refer to https://github.com/KosinskiLab/AlphaPulldownSnakemake/tree/main and snakemake documentation for more details.

   👀React with 👀1Reacted by MMO1313
2. MauriAndresMU1313 commented on Nov 21, 2025

   

   MauriAndresMU1313

   on Nov 21, 2025

   Author

   More actions

   Thank you for provide that feedback! My bad for the basic questions...

   Now looks like is finally working, at least in the dry-run option. I'm using the test data specified on the sample sheet file. In this case, I'm having the following error:

   /bin/bash: line 3: create_individual_features.py: command not found
   

   Click to expand

   /bin/bash: line 3: create_individual_features.py: command not found
   RuleException:
   CalledProcessError in file "https://raw.githubusercontent.com/KosinskiLab/AlphaPulldownSnakemake/2.1.5/workflow/Snakefile", line 205:
   Command ' singularity  exec --home '/scratch/mmora30/workspace/AlphaPulldownSnakemake' --nv --bind '/users/mmora30/.cache/snakemake/snakemake/source-cache' /users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/fold_analysis_latest.sif bash -c 'set -euo pipefail;
           create_individual_features.py \
               --fasta_paths=/users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/data/P01258.fasta \
               --data_dir=/users/mmora30/rmedin6-g00/full_db \
               --output_dir=/users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/features \
               --save_msa_files=False --use_precomputed_msas=True --max_template_date=2050-01-01 --compress_features=True --data_pipeline=alphafold2'' returned non-zero exit status 127.
     File "https://raw.githubusercontent.com/KosinskiLab/AlphaPulldownSnakemake/2.1.5/workflow/Snakefile", line 205, in __rule_create_features
   [Fri Nov 21 16:36:57 2025]
   Error in rule create_features:
       message: None
       jobid: 4
       input: /users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/data/P01258.fasta
       output: /users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/features/P01258.pkl.xz
       shell:
   
           create_individual_features.py \
               --fasta_paths=/users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/data/P01258.fasta \
               --data_dir=/users/mmora30/rmedin6-g00/full_db \
               --output_dir=/users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/features \
               --save_msa_files=False --use_precomputed_msas=True --max_template_date=2050-01-01 --compress_features=True --data_pipeline=alphafold2
   
           (command exited with non-zero exit code)
   Exiting because a job execution failed. Look below for error messages
   [Fri Nov 21 16:36:57 2025]
   Error in rule create_features:
       message: None
       jobid: 4
       input: /users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/data/P01258.fasta
       output: /users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/features/P01258.pkl.xz
       shell:
   
           create_individual_features.py \
               --fasta_paths=/users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/data/P01258.fasta \
               --data_dir=/users/mmora30/rmedin6-g00/full_db \
               --output_dir=/users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/features \
               --save_msa_files=False --use_precomputed_msas=True --max_template_date=2050-01-01 --compress_features=True --data_pipeline=alphafold2
   
           (command exited with non-zero exit code)
   [Fri Nov 21 16:36:57 2025]
   Error in rule create_features:
       message: None
       jobid: 6
       input: /users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/data/P01579.fasta
       output: /users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/features/P01579.pkl.xz
       shell:
   
           create_individual_features.py \
               --fasta_paths=/users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/data/P01579.fasta \
               --data_dir=/users/mmora30/rmedin6-g00/full_db \
               --output_dir=/users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs/features \
               --save_msa_files=False --use_precomputed_msas=True --max_template_date=2050-01-01 --compress_features=True --data_pipeline=alphafold2
   
           (command exited with non-zero exit code)
   Complete log(s): /scratch/mmora30/workspace/AlphaPulldownSnakemake/.snakemake/log/2025-11-21T163656.396070.snakemake.log
   WorkflowError:
   At least one job did not complete successfully.
   

   I did not find similar error from another user, apparently this line is critical:

   --save_msa_files=False --use_precomputed_msas=True --max_template_date=2050-01-01 --compress_features=True --data_pipeline=alphafold2'' returned non-zero exit status 127.
   

   Could you provide some feedback? Looks like is something related to MSA process, however, I checked the create_individual_features.py script, but in the line 3 I did not find relevant information.

   Do you need that I provide more information?

   Thank you again for your feedback and help!!
3. MauriAndresMU1313 commented on Nov 24, 2025

   

   MauriAndresMU1313

   on Nov 24, 2025

   Author

   More actions

   Looking forward to your feedback @DimaMolod
   Thank you again!
4. DimaMolod commented on Nov 24, 2025

   

   DimaMolod

   on Nov 24, 2025

   Collaborator

   More actions

   Hi @MauriAndresMU1313
   the problem that you mixed up containers in you config.yaml file and try to use container for analysis pipeline for predictions.
   Please check your config.yaml carefully and use the latest version.
   BR
   Dima
5. MauriAndresMU1313 commented on Nov 24, 2025

   

   MauriAndresMU1313

   on Nov 24, 2025

   Last edited by MauriAndresMU1313

   Author

   More actions

   Wow! Thank you for the quick response.
   Let me share with you the config.yaml file:

   # List of input sample sheets
   input_files:
     - config/sample_sheet.csv
   #  - config/sample_sheet.csv
   
   # Delimiter used in protein names
   protein_delimiter: "+"
   
   # Whether to drop equivalent permutations (AB and BA are the same) when generating fold specs
   exclude_permutations: true
   
   # Directory where all output files will be stored
   output_directory: /users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/outputs
   
   # Path to AlphaFold databases and to backend weights
   # Note prior to version 2.0.4 these were called alphafold_data_directory
   databases_directory: /users/mmora30/rmedin6-g00/full_db
   backend_weights_directory : /users/mmora30/rmedin6-g00/full_db/params
   
   # Directories containing precomputed features
   feature_directory:
     - "/path/to/directory/with/features1"
     - "/path/to/directory/with/features2"
   
   # If True, only generate features without running structure prediction
   only_generate_features: false
   
   # Toggle structure analysis after prediction
   enable_structure_analysis: true
   
   # Generate an aggregated CSV (reports/all_interfaces.csv) across all folds
   generate_recursive_report: true
   
   # Extra CLI flags to pass to alphajudge during recursive report generation
   recursive_report_arguments: {}
   
   # Arguments for feature generation
   create_feature_arguments:
     --save_msa_files: false  # Do not save multiple sequence alignment (MSA) files
     --use_precomputed_msas: true  # Use precomputed MSA files if available
     --max_template_date: 2050-01-01  # Set maximum template date to include all templates
     --compress_features: true  # Compress generated features
     --data_pipeline: alphafold2 # Use alphafold2 or alphafold3 data pipeline for generating features
   
   # Arguments for structure inference
   structure_inference_arguments:
     --fold_backend: alphafold3 # Options: alphafold2, alphafold3, alphalink
   
   # Arguments for structure analysis
   analyze_structure_arguments:
     --contact_thresh: 8.0 # Contact threshold for structure analysis
     --pae_filter: 100.0 # PAE filter for structure analysis
     --models_to_analyse: "all"  # Models to analyze
   
   # Memory allocation settings for feature creation and structure inference
   feature_create_ram_bytes: 64000
   feature_create_ram_scaling: 1.1
   structure_inference_ram_bytes: 64000
   
   # Number of threads for AlphaFold inference
   alphafold_inference_threads: 8
   
   # SLURM resources
   slurm_partition: "gpu"
   slurm_gres: "gpu:1"
   slurm_qos: "normal"
   
   # Specify the backend by changing the prediction container
   # (you can also use local singularity .sif files)
   # - "docker://kosinskilab/fold:2.1.2" for AlphaFold2
   # - "docker://kosinskilab/alphafold3:2.1.2" for AlphaFold3
   # - "docker://kosinskilab/alphalink:2.1.2" for AlphaLink2
   # - "/path/to/my/container.sif"
   prediction_container: "/users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/fold_analysis_latest.sif"
   
   #"docker://kosinskilab/alphafold3:latest"
   
   # Container for structure analysis
   analysis_container: "/users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/alphajudge.sif"
   

   So far I downloaded the latest containers, here you can see:

   fold_analysis_latest.sif
   -rwxr-xr-x.  1 mmora30 HyPER-LG-PROD-Sec 7667847168 Nov 21 00:57 fold_analysis_with_ccp4.sif
   drwxr-sr-x.  4 mmora30 HyPER-LG-PROD-Sec      25600 Nov 21 15:27 features_input
   -rwxr-xr-x.  1 mmora30 HyPER-LG-PROD-Sec  101879808 Nov 21 15:50 alphajudge.sif
   

   All the process to obtain the container:

   # Pull the Singularity image
   singularity pull docker://kosinskilab/fold_analysis:latest
   
   singularity build --sandbox ap_sandbox fold_analysis_latest.sif
   
   tar xvzf ccp4-9.0.011-linux64.tar.gz
   
   cp ccp4-9/bin/pisa ccp4-9/bin/sc /users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/ap_sandbox/software
   
   cp -rn lib/* /users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/ap_sandbox/software/lib64/
   
   singularity build fold_analysis_with_ccp4.sif /users/mmora30/mmora30/workspace/AlphaPulldownSnakemake/ap_sandbox
   
   singularity pull alphajudge.sif docker://kosinskilab/alphajudge:latest
   

   Could you provide some feedback to know which container could be the source of the issue?

   Do you see something unusual?

   Do you need that I provide more details to make the process smoothly?

   Thank you again!
6. KosinskiLab

   locked and limited conversation to collaborators on Nov 24, 2025
7. DimaMolod

   converted this issue into a discussion #559 on Nov 24, 2025