Code for training and testing PoinVS models for paper (preprint at doi.org/10.1101/2022.10.28.511712).
EGNN (paper) is an E(3)-equivariant graph neural network layer. In this project, we use networks based on this to make E(3)-invariant predictions of binding pose and affinity.
The author's previous work on deep learning for virtual screening can be found here.
Copy and paste the following commands:
ARCH=$(arch)
if [[ "$OSTYPE" == "linux-gnu"* && "$ARCH" == "x86_64"* ]]; then
INSTALLSCRIPT="environment.yml"
elif [[ "$OSTYPE" == "darwin"* && "$ARCH" == "arm"* ]]; then
INSTALLSCRIPT="environment_apple_silicon.yml"
elif [[ "$OSTYPE" == "darwin"* && "$ARCH" == "x86_64"* ]]; then
INSTALLSCRIPT="environment_macOS_intel.yml"
else
echo "$OSTYPE with $ARCH not supported. Abort..."
exit
fi
git remote set-url origin git@github.com:jscant/PointVS.git
conda env create -f $INSTALLSCRIPT
conda activate pointvs
pip install -e .
Note: Windows and Linux (arm64) are not supported. Apple Silicon devices will use MPS where available, although this can harm performance on smaller models.
To run tests for invariance under E(3) transformations:
python3 -m pytest -vvv
If you would like to use wandb for logging information on loss, performance and hyperparameters (recommended), you must first create an account at wandb.ai. You must then log into your account on your local machine by opening a python3 console and executing:
import wandb
wandb.init()
and following the onscreen instructions.
A small working example is:
python3 point_vs.py multitask /tmp/test_run --model_task both \
-ea 1 -ep 1 --layers 3 \
--train_data_root_pose data/small_chembl_test \
--train_types_pose data/small_chembl_test.types \
--train_data_root_affinity data/multi_classification_sample \
--train_types_affinity data/multi_classification_sample.types \
--test_data_root_pose data/small_chembl_test \
--test_types_pose data/small_chembl_test.types \
--test_types_affinity data/multi_classification_sample.types \
--test_data_root_affinity data/multi_classification_sample
usage: point_vs.py [-h] [--train_data_root_pose TRAIN_DATA_ROOT_POSE]
[--train_data_root_affinity TRAIN_DATA_ROOT_AFFINITY]
[--test_data_root_pose TEST_DATA_ROOT_POSE]
[--test_data_root_affinity TEST_DATA_ROOT_AFFINITY]
[--logging_level LOGGING_LEVEL] [--load_weights LOAD_WEIGHTS]
[--test_data_root TEST_DATA_ROOT]
[--translated_actives TRANSLATED_ACTIVES]
[--batch_size BATCH_SIZE] [--epochs_pose EPOCHS_POSE]
[--epochs_affinity EPOCHS_AFFINITY] [--channels CHANNELS]
[--learning_rate LEARNING_RATE] [--weight_decay WEIGHT_DECAY]
[--wandb_project WANDB_PROJECT] [--wandb_run WANDB_RUN]
[--layers LAYERS] [--radius RADIUS] [--load_args LOAD_ARGS]
[--double] [--activation ACTIVATION] [--dropout DROPOUT]
[--use_1cycle] [--warm_restarts]
[--fourier_features FOURIER_FEATURES] [--norm_coords]
[--norm_feats] [--use_atomic_numbers] [--compact] [--thin_mlps]
[--hydrogens] [--augmented_actives AUGMENTED_ACTIVES]
[--min_aug_angle MIN_AUG_ANGLE]
[--max_active_rmsd MAX_ACTIVE_RMSD]
[--min_inactive_rmsd MIN_INACTIVE_RMSD] [--val_on_epoch_end]
[--synth_pharm] [--input_suffix INPUT_SUFFIX]
[--train_types_pose TRAIN_TYPES_POSE]
[--train_types_affinity TRAIN_TYPES_AFFINITY]
[--test_types_pose TEST_TYPES_POSE]
[--test_types_affinity TEST_TYPES_AFFINITY] [--egnn_attention]
[--egnn_tanh] [--egnn_normalise] [--egnn_residual]
[--edge_radius EDGE_RADIUS] [--end_flag] [--wandb_dir WANDB_DIR]
[--estimate_bonds] [--prune] [--top1] [--graphnorm]
[--multi_fc] [--lucid_node_final_act]
[--p_remove_entity P_REMOVE_ENTITY] [--static_coords]
[--permutation_invariance] [--node_attention]
[--attention_activation_function ATTENTION_ACTIVATION_FUNCTION]
[--node_attention_final_only] [--edge_attention_final_only]
[--node_attention_first_only] [--edge_attention_first_only]
[--only_save_best_models] [--egnn_edge_residual]
[--gated_residual] [--rezero] [--extended_atom_types]
[--max_inactive_rmsd MAX_INACTIVE_RMSD] [--model_task MODEL_TASK]
[--synthpharm] [--p_noise P_NOISE] [--include_strain_info]
[--final_softplus] [--optimiser OPTIMISER]
[--multi_target_affinity]
model save_path
positional arguments:
model Type of point cloud network to use: lucid or egnn
save_path Directory in which experiment outputs are stored. If
wandb_run and wandb_project are specified,
save_path/wandb_project/wandb_run will be used to store
results.
optional arguments:
-h, --help show this help message and exit
--train_data_root_pose TRAIN_DATA_ROOT_POSE
Location relative to which parquets files for training the
pose classifier as specified in the train_types_pose file are
stored.
--train_data_root_affinity TRAIN_DATA_ROOT_AFFINITY, --tdra TRAIN_DATA_ROOT_AFFINITY
Location relative to which parquets files for training the
affinity predictor as specified in the train_types file are
stored.
--test_data_root_pose TEST_DATA_ROOT_POSE
Location relative to which parquets files for testing the
pose classifier as specified in the test_types_pose file are
stored.
--test_data_root_affinity TEST_DATA_ROOT_AFFINITY
Location relative to which parquets files for testing the
affinity predictor as specified in the test_types file are
stored.
--logging_level LOGGING_LEVEL
Level at which to print logging statements. Any of notset,
debug, info, warning, error, critical.
--load_weights LOAD_WEIGHTS, -l LOAD_WEIGHTS
Load a model.
--test_data_root TEST_DATA_ROOT, -t TEST_DATA_ROOT
Location of structure test *.parquets files. Receptors should
be in a directory named receptors, with ligands located in
their specific receptor subdirectory under the ligands
directory.
--translated_actives TRANSLATED_ACTIVES
Directory in which translated actives are stored. If
unspecified, no translated actives will be used. The use of
translated actives are is discussed in
https://pubs.acs.org/doi/10.1021/acs.jcim.0c00263
--batch_size BATCH_SIZE, -b BATCH_SIZE
Number of examples to include in each batch for training.
--epochs_pose EPOCHS_POSE, -ep EPOCHS_POSE
Number of times to iterate through pose training set.
--epochs_affinity EPOCHS_AFFINITY, -ea EPOCHS_AFFINITY
Number of times to iterate through affinity training set.
--channels CHANNELS, -k CHANNELS
Channels for feature vectors
--learning_rate LEARNING_RATE, -lr LEARNING_RATE
Learning rate for gradient descent
--weight_decay WEIGHT_DECAY, -w WEIGHT_DECAY
Weight decay for regularisation
--wandb_project WANDB_PROJECT
Name of wandb project. If left blank, wandb logging will not
be used.
--wandb_run WANDB_RUN
Name of run for wandb logging.
--layers LAYERS Number of layers in LieResNet
--radius RADIUS Maximum distance from a ligand atom for a receptor atom to be
included in input
--load_args LOAD_ARGS
Load yaml file with command line args. Any args specified in
the file will overwrite other args specified on the command
line.
--double Use 64-bit floating point precision
--activation ACTIVATION
Activation function
--dropout DROPOUT Chance for nodes to be inactivated on each trainin batch
(EGNN)
--use_1cycle Use 1cycle learning rate scheduling
--warm_restarts Use cosine annealing with warm restarts
--fourier_features FOURIER_FEATURES
(Lucid) Number of fourier terms to use when encoding
distances (default is not to use fourier distance encoding)
--norm_coords (Lucid) Normalise coordinate vectors
--norm_feats (Lucid) Normalise feature vectors
--use_atomic_numbers Use atomic numbers rather than smina types
--compact Use compact rather than true one-hot encodings
--thin_mlps (Lucid) Use single layer MLPs for edge, node and coord
updates
--hydrogens Include polar hydrogens
--augmented_actives AUGMENTED_ACTIVES
Number of randomly rotated actives to be included as decoys
during training
--min_aug_angle MIN_AUG_ANGLE
Minimum angle of rotation for augmented actives as specified
in the augmented_actives argument
--max_active_rmsd MAX_ACTIVE_RMSD
(Pose selection) maximum non-aligned RMSD between the
original crystal pose and active redocked poses
--min_inactive_rmsd MIN_INACTIVE_RMSD
(Pose selection) minimum non-aligned RMSD between original
crystal pose and inactive redocked poses
--val_on_epoch_end, -v
Run inference ion the validation set at the end of every
epoch during training
--synth_pharm, -p Synthetic Pharmacophore mode (for Tom, beta)
--input_suffix INPUT_SUFFIX, -s INPUT_SUFFIX
Filename extension for inputs
--train_types_pose TRAIN_TYPES_POSE
Optional name of GNINA-like types file which contains paths
and labels for a pose training set. See GNINA 1.0
documentation for specification.
--train_types_affinity TRAIN_TYPES_AFFINITY
Optional name of GNINA-like types file which contains paths
and labels for an affinity training set. See GNINA 1.0
documentation for specification.
--test_types_pose TEST_TYPES_POSE
Optional name of GNINA-like types file which contains paths
and labels for a pose test set. See GNINA 1.0 documentation
for specification.
--test_types_affinity TEST_TYPES_AFFINITY
Optional name of GNINA-like types file which contains paths
and labels for an affinity test set. See GNINA 1.0
documentation for specification.
--egnn_attention Use attention mechanism on edges for EGNN
--egnn_tanh Put tanh layer at the end of the coordinates mlp (EGNN)
--egnn_normalise Normalise radial coordinates (EGNN)
--egnn_residual Use residual connections (EGNN)
--edge_radius EDGE_RADIUS
Maximum interatomic distance for an edge to exist (EGNN)
--end_flag Add a file named "_FINISHED" to the save_path upon training
and test completion
--wandb_dir WANDB_DIR
Location to store wandb files. Defaults to
<save_path>/<wandb_project>/<wandb_run>/wandb.
--estimate_bonds (EGNN): Instead of using a fixed edge radius,the
intermolecular radius is set at --edge_radius Angstroms but
the intramolecular radius is set at 2A, which has the effect
of putting edges where there are covalent bonds between atoms
in the same molecule.
--prune (EGNN) Prune subgraphs which are not connected to the ligand
--top1 A poorly kept secret ;)
--graphnorm (EGNN) add GraphNorm layers to each node MLP
--multi_fc Three fully connected layers rather than just one to
summarise the graph at the end of the EGNN
--lucid_node_final_act
(Lucid) SiLU at the end of node MLPs
--p_remove_entity P_REMOVE_ENTITY
Rate at which one of (randomly selected) ligand or receptor
is removed and label is forced to zero
--static_coords Do not update coords (eq. 4, EGNN)
--permutation_invariance
Edge features are invariant to order of input node (EGNN,
experimental)
--node_attention Use attention mechanism for nodes
--attention_activation_function ATTENTION_ACTIVATION_FUNCTION
One of sigmoid, relu, silu or tanh
--node_attention_final_only
Only apply attention mechanism to nodes in the final layer
--edge_attention_final_only
Only apply attention mechanism to edges in the final layer
--node_attention_first_only
Only apply attention mechanism to nodes in the first layer
--edge_attention_first_only
Only apply attention mechanism to edges in the first layer
--only_save_best_models
Only save models which improve upon previous models
--egnn_edge_residual Residual connections for individual messages (EGNN)
--gated_residual Residual connections are gated by a single learnable
parameter (EGNN), see
home.ttic.edu/~savarese/savarese_files/Residual_Gates.pdf
--rezero ReZero residual connections (EGNN), see
arxiv.org/pdf/2003.04887.pdf
--extended_atom_types
18 atom types rather than 10
--max_inactive_rmsd MAX_INACTIVE_RMSD
Discard structures beyond <x> RMSD from xtal pose
--model_task MODEL_TASK
One of either classification or regression;
--synthpharm For tom
--p_noise P_NOISE Probability of label being inverted during training
--include_strain_info
Include info on strain energy and RMSD from ground state of
ligand
--final_softplus Final layer in regression has softplus nonlinearity
--optimiser OPTIMISER, -o OPTIMISER
Optimiser (either adam or sgd)
--multi_target_affinity
Use multitarget regression for affinity. If True, targets are
split depending on if labels are pkd, pki or IC50.
For full instructions, see the README.
The input format for PointVS neural networks is the
parquet. The script
point_vs/dataset_generation/generate_types_file.py can be used to generate
the types file (Loading Datasets section below), from which
point_vs/dataset_generation/types_to_parquet.py can be used to convert
pdb, sdf and mol2 files to the required format.
The recommended way to feed data into the models is to use types files, which are used extensively in GNINA 1.0.
Types File Each line should follow one of two formats.
For classification datasets:
<label> <energy> <rmsd> <receptor> <ligand>
For regression datasets:
<pki> <pkd> <ic50> <receptor> <ligand>
where:
<label> is a binary label (0 or 1)
<energy> is usually set to -1.0, and can be used to store the energy of
an interaction or a docking score. It is included in the original GNINA
specification so is retained here.
<rmsd> is the RMSD from the original crystal structure where the ligand is a
docked structure and the RMSD is known, or -1 otherwise.
<pki> The pKi, if known, else -1.0
<pkd> The pKd, if known, else -1.0
<ic50> The IC50, if known, else -1.0.
<receptor> is the relative location of the receptor structure in parquet format
<ligand> is the relative location of the ligand structure in parquet format
Only one of , or should have a positive value. If multiple types of affinity are known, they can be given separate entries.
An example
types file can be found at data/small_chembl_test.types, where the <rmsd>
value is not known:
1 -1 -1.0 receptors/12968.parquet ligands/12968_actives/mol25_7.parquet
0 -1 -1.0 receptors/12968.parquet ligands/12968_decoys/mol9690_0.parquet
0 -1 -1.0 receptors/12968.parquet ligands/12968_decoys/mol6981_5.parquet
0 -1 -1.0 receptors/12968.parquet ligands/12968_decoys/mol2542_5.parquet
See the README
in point_vs/dataset_generation for a script which generates types files and
calculates RMSDs.