Merge pull request #7 from BioinfoMachineLearning/af

v0.5.0 - Add results with AlphaFold 3 predicted structures and for Chai-1

.github/workflows/code-quality-main.yaml

@@ -21,4 +21,4 @@ jobs:
           python-version: "3.10"
 
       - name: Run pre-commits
-        uses: pre-commit/action@v2.0.3
+        uses: pre-commit/action@v3.0.1

.github/workflows/code-quality-pr.yaml

@@ -33,6 +33,6 @@ jobs:
         run: echo '${{ steps.file_changes.outputs.files}}'
 
       - name: Run pre-commits
-        uses: pre-commit/action@v2.0.3
+        uses: pre-commit/action@v3.0.1
         with:
           extra_args: --files ${{ steps.file_changes.outputs.files}}

.github/workflows/release-drafter.yml

@@ -22,6 +22,6 @@ jobs:
 
     steps:
       # Drafts your next Release notes as Pull Requests are merged into "master"
-      - uses: release-drafter/release-drafter@v5
+      - uses: release-drafter/release-drafter@v6
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.gitignore

@@ -163,6 +163,9 @@ configs/local/default.yaml
 
 # Forks
 /workdir/
+/forks/chai-lab/chai-lab/
+/forks/chai-lab/prediction_inputs/
+/forks/chai-lab/prediction_outputs/
 /forks/DiffDock1.0/
 /forks/DiffDock/DiffDock/
 /forks/DynamicBind/*.npy
@@ -189,4 +192,4 @@ configs/local/default.yaml
 /forks/RoseTTAFold-All-Atom/psipred/
 /forks/TULIP/outputs/
 /forks/Vina/ADFR/
-scripts/inference/
+scripts/*inference*/

CHANGELOG.md

@@ -1,3 +1,12 @@
+### 0.5.0 - 09/30/2024
+
+- Added results with AlphaFold 3 predicted structures (now the default)
+- Added results for the new Chai-1 model from Chai Discovery
+- Added a new inference sweep pipeline for HPC clusters to allow users to quickly run an exhaustive sweep of all baseline methods, datasets, and tasks e.g., using generated batch scripts and a SLURM scheduler
+- Updated Zenodo links to point to the latest version of the project's Zenodo record, which now includes the above-mentioned AlphaFold 3 predicted structures and baseline method results using them
+- Updated documentation project-wide according to the additions listed above
+- Fixed some CI testing issues
+
 ### 0.4.0 - 08/12/2024
 
 - Renamed `src` root directory to `posebench` to support `pip` packaging

configs/analysis/complex_alignment.yaml

@@ -1,11 +1,12 @@
-method: neuralplexer # the method for which to align predictions - NOTE: must be one of (`diffdock`, `fabind`, `dynamicbind`, `neuralplexer`, `rfaa`)
-vina_binding_site_method: diffdock # the method to use for Vina binding site prediction - NOTE: must be one of (`diffdock`, `fabind`, `dynamicbind`, `neuralplexer`, `rfaa`)
+method: neuralplexer # the method for which to align predictions - NOTE: must be one of (`diffdock`, `fabind`, `dynamicbind`, `neuralplexer`, `rfaa`, `chai-lab`)
+vina_binding_site_method: diffdock # the method to use for Vina binding site prediction - NOTE: must be one of (`diffdock`, `fabind`, `dynamicbind`, `neuralplexer`, `rfaa`, `chai-lab`)
 dataset: posebusters_benchmark # the dataset to use - NOTE: must be one of (`posebusters_benchmark`, `astex_diverse`, `dockgen`, `casp15`)
 ensemble_ranking_method: consensus # the method with which to rank-order and select the top ensemble prediction for each target - NOTE: must be one of (`consensus`, `ff`)
 input_data_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set # the input protein-ligand complex directory to recursively parse
-output_dir: ${resolve_method_output_dir:${method},${dataset},${vina_binding_site_method},${ensemble_ranking_method},${repeat_index}} # the output directory to which to save the relaxed predictions
+output_dir: ${resolve_method_output_dir:${method},${dataset},${vina_binding_site_method},${ensemble_ranking_method},${repeat_index},${pocket_only_baseline},${v1_baseline}} # the output directory to which to save the relaxed predictions
 rank_to_align: 1 # the pose rank to align
-aligned_filename_postfix: "_aligned" # the postfix to append to each aligned complex filename
+aligned_filename_suffix: "_aligned" # the suffix to append to each aligned complex filename
 force_process: false # whether to force processing of all complexes, even if they have already been processed
 repeat_index: 1 # the repeat index which was used for inference
 pocket_only_baseline: false # whether to prepare the pocket-only baseline
+v1_baseline: false # whether to prepare the v1 baseline

configs/analysis/inference_analysis.yaml

@@ -1,13 +1,14 @@
 full_report: true # whether to generate a full PoseBusters report (i.e. with all metrics) or a summary report (i.e. with only the most important metrics)
-method: diffdock # the method for which to score predictions - NOTE: must be one of (`diffdock`, `fabind`, `dynamicbind`, `neuralplexer`, `rfaa`, `vina`, `ensemble`)
-vina_binding_site_method: diffdock # the method to use for Vina binding site prediction - NOTE: must be one of (`diffdock`, `fabind`, `dynamicbind`, `neuralplexer`, `rfaa`, `p2rank`)
+method: diffdock # the method for which to score predictions - NOTE: must be one of (`diffdock`, `fabind`, `dynamicbind`, `neuralplexer`, `rfaa`, `chai-lab`, `vina`, `ensemble`)
+vina_binding_site_method: diffdock # the method to use for Vina binding site prediction - NOTE: must be one of (`diffdock`, `fabind`, `dynamicbind`, `neuralplexer`, `rfaa`, `chai-lab`, `p2rank`)
 dataset: posebusters_benchmark # the dataset to use - NOTE: must be one of (`posebusters_benchmark`, `astex_diverse`, `dockgen`, `casp15`)
 ensemble_ranking_method: consensus # the method with which to rank-order and select the top ensemble prediction for each target - NOTE: must be one of (`consensus`, `ff`)
-input_csv_path: ${resolve_method_input_csv_path:${method},${dataset}} # the input CSV filepath with which to run inference
+input_csv_path: ${resolve_method_input_csv_path:${method},${dataset},${pocket_only_baseline}} # the input CSV filepath with which to run inference
 input_data_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set # the input protein-ligand complex directory to recursively parse
 posebusters_ccd_ids_filepath: ${oc.env:PROJECT_ROOT}/data/posebusters_pdb_ccd_ids.txt # the path to the PoseBusters PDB CCD IDs file that lists the targets that do not contain any crystal contacts
-dockgen_test_ids_filepath: ${oc.env:PROJECT_ROOT}/data/dockgen_set/split_test_rmsd_filtered.txt # the path to the (ESMFold RMSD-filtered) DockGen test set IDs file
-output_dir: ${resolve_method_output_dir:${method},${dataset},${vina_binding_site_method},${ensemble_ranking_method},${repeat_index}} # the output directory to which to save the relaxed predictions
+dockgen_test_ids_filepath: ${oc.env:PROJECT_ROOT}/data/dockgen_set/split_test_rmsd_filtered.txt # the path to the (predicted RMSD-filtered) DockGen test set IDs file
+output_dir: ${resolve_method_output_dir:${method},${dataset},${vina_binding_site_method},${ensemble_ranking_method},${repeat_index},${pocket_only_baseline},${v1_baseline}} # the output directory to which to save the relaxed predictions
 repeat_index: 1 # the repeat index which was used for inference
 pocket_only_baseline: false # whether to analyze the pocket-only baseline
+v1_baseline: false # whether to analyze the v1 baseline
 relax_protein: false # whether to relax the protein - NOTE: currently periodically yields unpredictable protein-ligand separation

configs/analysis/inference_analysis_casp.yaml

@@ -1,8 +1,8 @@
 full_report: true # whether to generate a full PoseBusters report (i.e. with all metrics) or a summary report (i.e. with only the most important metrics)
 python_exec_path: ${oc.env:HOME}/mambaforge/envs/casp15_ligand_scoring/bin/python3 # the Python executable to use
 scoring_script_path: ${oc.env:PROJECT_ROOT}/posebench/analysis/casp15_ligand_scoring/score_predictions.py # the path to the script to use for scoring CASP predictions
-method: diffdock # the method for which to score predictions - NOTE: must be one of (`diffdock`, `fabind`, `dynamicbind`, `neuralplexer`, `rfaa`, `vina`, `ensemble`, `tulip`)
-vina_binding_site_method: diffdock # the method to use for Vina binding site prediction - NOTE: must be one of (`diffdock`, `fabind`, `dynamicbind`, `neuralplexer`, `rfaa`)
+method: diffdock # the method for which to score predictions - NOTE: must be one of (`diffdock`, `fabind`, `dynamicbind`, `neuralplexer`, `rfaa`, `chai-lab`, `vina`, `ensemble`, `tulip`)
+vina_binding_site_method: diffdock # the method to use for Vina binding site prediction - NOTE: must be one of (`diffdock`, `fabind`, `dynamicbind`, `neuralplexer`, `rfaa`, `chai-lab`)
 dataset: casp15 # the dataset to use - NOTE: must be one of (`casp15`)
 ensemble_ranking_method: consensus # the method with which to rank-order and select the top ensemble prediction for each target - NOTE: must be one of (`consensus`, `ff`)
 predictions_dir: ${oc.env:PROJECT_ROOT}/data/test_cases/${dataset}/top_${method}_ensemble_predictions_${repeat_index} # the directory containing the predictions to analyze
@@ -12,4 +12,6 @@ fault_tolerant: true # whether to continue processing targets if an error occurs
 skip_existing: true # whether to skip processing targets for which output already exists
 score_relaxed_structures: true # whether to score relaxed structures in addition to the original (unrelaxed) structures
 repeat_index: 1 # the run index to use for scoring predictions
-no_pretraining: false # whether to score a model without pretraining
+no_ilcl: false # whether to score a model trained without an inter-ligand clash loss (ILCL) - NOTE: only applicable to the `neuralplexer` method
+relax_protein: false # whether to relax the protein - NOTE: currently periodically yields unpredictable protein-ligand separation
+v1_baseline: false # whether to score the v1 baseline predictions

configs/data/chai_input_preparation.yaml

@@ -0,0 +1,9 @@
+dataset: posebusters_benchmark # the dataset to use - NOTE: must be one of (`posebusters_benchmark`, `astex_diverse`, `dockgen`, `casp15`)
+input_data_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set # the input protein-ligand complex directory to recursively parse
+output_scripts_path: ${oc.env:PROJECT_ROOT}/forks/chai-lab/prediction_inputs/${dataset} # the output directory in which to save the input files
+posebusters_ccd_ids_filepath: ${oc.env:PROJECT_ROOT}/data/posebusters_pdb_ccd_ids.txt # the path to the PoseBusters PDB CCD IDs file that lists the targets that do not contain any crystal contacts
+dockgen_test_ids_filepath: ${oc.env:PROJECT_ROOT}/data/dockgen_set/split_test.txt # the path to the DockGen test set IDs file
+protein_filepath: null # the path to the protein structure file to use
+ligand_smiles: null # the ligand SMILES string for which to predict the binding pose
+input_id: null # the input ID to use for inference
+pocket_only_baseline: false # whether to prepare the pocket-only baseline

configs/data/chai_output_extraction.yaml

@@ -0,0 +1,13 @@
+dataset: posebusters_benchmark # the dataset to use - NOTE: must be one of (`posebusters_benchmark`, `astex_diverse`, `dockgen`, `casp15`)
+prediction_inputs_dir: ${oc.env:PROJECT_ROOT}/forks/chai-lab/prediction_inputs/${dataset}
+prediction_outputs_dir: ${oc.env:PROJECT_ROOT}/forks/chai-lab/prediction_outputs/${dataset}_${repeat_index}
+inference_outputs_dir: ${oc.env:PROJECT_ROOT}/forks/chai-lab/inference/chai-lab_${dataset}_outputs_${repeat_index}
+input_data_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set # the input protein-ligand complex directory to recursively parse
+posebusters_ccd_ids_filepath: ${oc.env:PROJECT_ROOT}/data/posebusters_pdb_ccd_ids.txt # the path to the PoseBusters PDB CCD IDs file that lists the targets that do not contain any crystal contacts
+dockgen_test_ids_filepath: ${oc.env:PROJECT_ROOT}/data/dockgen_set/split_test.txt # the path to the DockGen test set IDs file
+complex_filepath: null # if not `null`, this should be the path to the complex PDB file for which to extract outputs
+complex_id: null # if not `null`, this should be the complex ID of the single complex for which to extract outputs
+ligand_smiles: null # if not `null`, this should be the (i.e., `.` fragment-separated) complex ligand SMILES string of the single complex for which to extract outputs
+output_dir: null # if not `null`, this should be the path to the output file to which to write the extracted outputs
+repeat_index: 1 # the repeat index with which inference was run
+pocket_only_baseline: false # whether to prepare the pocket-only baseline

configs/data/diffdock_input_preparation.yaml

@@ -1,6 +1,6 @@
 dataset: posebusters_benchmark # the dataset to use - NOTE: must be one of (`posebusters_benchmark`, `astex_diverse`, `dockgen`, `casp15`)
 input_data_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set # the input protein-ligand complex directory to recursively parse
-input_protein_structure_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set/${dataset}_holo_aligned_esmfold_structures # the input protein structure directory to parse
+input_protein_structure_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set/${dataset}_holo_aligned_predicted_structures # the input protein structure directory to parse
 output_csv_path: ${oc.env:PROJECT_ROOT}/forks/DiffDock/inference/diffdock_${dataset}_inputs.csv # the output CSV filepath to which to write the parsed input data
 posebusters_ccd_ids_filepath: ${oc.env:PROJECT_ROOT}/data/posebusters_pdb_ccd_ids.txt # the path to the PoseBusters PDB CCD IDs file that lists the targets that do not contain any crystal contacts
 dockgen_test_ids_filepath: ${oc.env:PROJECT_ROOT}/data/dockgen_set/split_test.txt # the path to the DockGen test set IDs file

configs/data/dynamicbind_input_preparation.yaml

@@ -6,3 +6,4 @@ posebusters_ccd_ids_filepath: ${oc.env:PROJECT_ROOT}/data/posebusters_pdb_ccd_id
 dockgen_test_ids_filepath: ${oc.env:PROJECT_ROOT}/data/dockgen_set/split_test.txt # the path to the DockGen test set IDs file
 protein_filepath: null # the path to the protein structure file to use
 ligand_smiles: null # the ligand SMILES string for which to predict the binding pose
+pocket_only_baseline: false # whether to prepare the pocket-only baseline

configs/data/fabind_input_preparation.yaml

@@ -3,3 +3,4 @@ input_data_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set # the input protein-l
 output_csv_path: ${oc.env:PROJECT_ROOT}/forks/FABind/inference/fabind_${dataset}_inputs.csv # the output CSV filepath to which to write the parsed input data
 posebusters_ccd_ids_filepath: ${oc.env:PROJECT_ROOT}/data/posebusters_pdb_ccd_ids.txt # the path to the PoseBusters PDB CCD IDs file that lists the targets that do not contain any crystal contacts
 dockgen_test_ids_filepath: ${oc.env:PROJECT_ROOT}/data/dockgen_set/split_test.txt # the path to the DockGen test set IDs file
+pocket_only_baseline: false # whether to prepare the pocket-only baseline

configs/data/neuralplexer_input_preparation.yaml

@@ -1,6 +1,6 @@
 dataset: posebusters_benchmark # the dataset to use - NOTE: must be one of (`posebusters_benchmark`, `astex_diverse`, `dockgen`, `casp15`)
 input_data_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set # the input protein-ligand complex directory to recursively parse
-input_receptor_structure_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set/${dataset}_holo_aligned_esmfold_structures # if not `null`, the input template protein structure directory to parse
+input_receptor_structure_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set/${dataset}_holo_aligned_predicted_structures # if not `null`, the input template protein structure directory to parse
 output_csv_path: ${oc.env:PROJECT_ROOT}/forks/NeuralPLexer/inference/neuralplexer_${dataset}_inputs.csv # the output CSV filepath to which to write the parsed input data
 posebusters_ccd_ids_filepath: ${oc.env:PROJECT_ROOT}/data/posebusters_pdb_ccd_ids.txt # the path to the PoseBusters PDB CCD IDs file that lists the targets that do not contain any crystal contacts
 dockgen_test_ids_filepath: ${oc.env:PROJECT_ROOT}/data/dockgen_set/split_test.txt # the path to the DockGen test set IDs file

configs/data/rfaa_output_extraction.yaml

@@ -7,3 +7,4 @@ complex_id: null # if not `null`, this should be the complex ID of the single co
 ligand_smiles: null # if not `null`, this should be the (i.e., `.` fragment-separated) complex ligand SMILES string of the single complex for which to extract outputs
 output_dir: null # if not `null`, this should be the path to the output file to which to write the extracted outputs
 repeat_index: 1 # the repeat index with which inference was run
+pocket_only_baseline: false # whether to prepare the pocket-only baseline

configs/data/tulip_output_extraction.yaml

@@ -2,5 +2,6 @@ dataset: posebusters_benchmark # the dataset to use - NOTE: must be one of (`pos
 prediction_outputs_dir: ${oc.env:PROJECT_ROOT}/forks/TULIP/outputs/${dataset}_${repeat_index}
 inference_outputs_dir: ${oc.env:PROJECT_ROOT}/forks/TULIP/inference/tulip_${dataset}_outputs_${repeat_index}
 posebusters_ccd_ids_filepath: ${oc.env:PROJECT_ROOT}/data/posebusters_pdb_ccd_ids.txt # the path to the PoseBusters PDB CCD IDs file that lists the targets that do not contain any crystal contacts
+dockgen_test_ids_filepath: ${oc.env:PROJECT_ROOT}/data/dockgen_set/split_test.txt # the path to the DockGen test set IDs file
 method_top_n_to_select: 5 # the number of top models for each target to select for analysis
 repeat_index: 1 # the repeat index to use

configs/model/chai_inference.yaml

@@ -0,0 +1,8 @@
+dataset: posebusters_benchmark # the dataset to use - NOTE: must be one of (`posebusters_benchmark`, `astex_diverse`, `dockgen`, `casp15`)
+input_dir: ${oc.env:PROJECT_ROOT}/forks/chai-lab/prediction_inputs/${dataset} # the input directory with which to run inference
+output_dir: ${oc.env:PROJECT_ROOT}/forks/chai-lab/prediction_outputs/${dataset}_${repeat_index} # the output directory to which to save the inference results
+cuda_device_index: 0 # the CUDA device to use for inference, or `null` to use CPU
+repeat_index: 1 # the repeat index to use for inference
+skip_existing: true # whether to skip running inference if the prediction for a target already exists
+pocket_only_baseline: false # whether to run the pocket-only baseline
+max_num_inputs: null # if provided, the number of (dataset subset) inputs over which to run inference

configs/model/diffdock_inference.yaml

@@ -14,4 +14,6 @@ actual_steps: 19 # the actual number of inference steps to run (i.e., after how
 no_final_step_noise: true # whether to disable the final inference step's noise from being added
 repeat_index: 1 # the repeat index to use for inference
 skip_existing: true # whether to skip inference for existing output directories
+pocket_only_baseline: false # whether to run the pocket-only baseline
 max_num_inputs: null # if provided, the number of (dataset subset) inputs over which to run inference
+v1_baseline: false # whether to run the v1 baseline

configs/model/dynamicbind_inference.yaml

@@ -2,12 +2,13 @@ cuda_device_index: 0 # the CUDA device to use for inference, or `null` to use CP
 python_exec_path: ${oc.env:PROJECT_ROOT}/forks/DynamicBind/DynamicBind/bin/python3 # the Python executable to use
 dynamicbind_exec_dir: ${oc.env:PROJECT_ROOT}/forks/DynamicBind # the DynamicBind directory in which to execute the inference scripts
 dataset: posebusters_benchmark # the dataset to use - NOTE: must be one of (`posebusters_benchmark`, `astex_diverse`, `dockgen`, `casp15`)
-input_data_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set/${dataset}_holo_aligned_esmfold_structures # the input protein-ligand complex directory to recursively parse for protein inputs
+input_data_dir: ${oc.env:PROJECT_ROOT}/data/${dataset}_set/${dataset}_holo_aligned_predicted_structures # the input protein-ligand complex directory to recursively parse for protein inputs
 input_ligand_csv_dir: ${oc.env:PROJECT_ROOT}/forks/DynamicBind/inference/dynamicbind_${dataset}_inputs # the input CSV directory with which to run inference
 samples_per_complex: 40 # the number of samples to generate per complex
 savings_per_complex: 1 # the (top-N) number of sample visualizations to save per complex
 inference_steps: 20 # the number of inference steps to run for each complex
 batch_size: 5 # the batch size to use for inference
+cache_path: ${oc.env:PROJECT_ROOT}/data/dynamicbind_cache/cache # the cache directory to use for storing intermediate data files
 header: ${dataset} # name of the results directory to create
 num_workers: 1 # the number of workers to use for native relaxation during inference
 skip_existing: true # whether to skip existing predictions