#!/bin/bash cat << 'EOF' ____ _ _ _ _ _ _ ____ _ _____ ____ _ _ ____ _ _ ____ _ ____ _ _ ____ __ ____ _ ____ _ __ __ _ / ___| _ _ _ __ / _| _ \ ___ ___| | __ | __ ) ___| |_ __ _ \ \ / /__ _ __ ___(_) ___ _ __ \___ \| | | | '__| |_| | | |/ _ \ / __| |/ / | _ \ / _ \ __/ _` | \ \ / / _ \ '__/ __| |/ _ \| '_ \ ___) | |_| | | | _| |_| | (_) | (__| < | |_) | __/ || (_| | \ V / __/ | \__ \ | (_) | | | | |____/ \__,_|_| |_| |____/ \___/ \___|_|\_\ |____/ \___|\__\__,_| \_/ \___|_| |___/_|\___/|_| |_| ____ _ _ _ _ _ _ ____ _ _____ ____ _ _ ____ _ _ ____ _ ____ _ _ EOF # This script is used to run SurfDock on test samples source ~/miniforge3/bin/activate SurfDock path=$(readlink -f "$0") SurfDockdir="$(dirname "$(dirname "$(dirname "$(dirname "$path")")")")" SurfDockdir=${SurfDockdir} echo SurfDockdir : ${SurfDockdir} temp="$(dirname "${SurfDockdir}")" model_temp=${SurfDockdir} #------------------------------------------------------------------------------------------------# #------------------------------------ Step1 : Setup Params --------------------------------------# #------------------------------------------------------------------------------------------------# export precomputed_arrays="${temp}/precomputed/precomputed_arrays" ## Please set the GPU devices you want to use gpu_string="7" echo "Using GPU devices: ${gpu_string}" IFS=',' read -ra gpu_array <<< "$gpu_string" NUM_GPUS=${#gpu_array[@]} export CUDA_VISIBLE_DEVICES=${gpu_string} ## Please set the main Parameters main_process_port=2957${gpu_array[-1]} ## Please set the project name project_name='SurfDock_Screen_samples_skip_target_processed' # /home/caoduanhua/NM_submit_code/SurfDock # Set default value for target_have_processed if not already set target_have_processed=${target_have_processed:-true} ## Please set the path to save the surface file and pocket file surface_out_dir=${temp}/Screen_result/processed_data/${project_name}/test_samples_8A_surface ## Please set the path to the input data data_dir=${SurfDockdir}/model/data/Screen_sample_dirs/test_samples ## Please set the path to the output csv file out_csv_dir=${temp}/Screen_result/processed_data/${project_name}/input_csv_files/ out_csv_file=${out_csv_dir}/test_samples.csv ## Please set the path to the esmbedding file esmbedding_dir=${temp}/Screen_result/processed_data/${project_name}/test_samples_esmbedding ## Please set the path to the Screen ligand library file Screen_lib_path=${SurfDockdir}/model/data/Screen_sample_dirs/test_samples/1a0q/1a0q_ligand_for_Screen.sdf ## Please set the path to the docking result directory docking_out_dir=${temp}/Screen_result/docking_result/${project_name} #------------------------------------------------------------------------------------------------# # -----------------------Step1 : Processed Target Structure -------------------------------------# #----------------(Set target_have_processed as true if you have done with your pipeline)---------# #------------------------------------------------------------------------------------------------# mkdir -p $surface_out_dir if [ "$target_have_processed" = true ]; then echo "Target structure has been processed, skipping this step." else echo "Processing target structure with OpenBabel..." export BABEL_LIBDIR=~/miniforge3/envs/SurfDock/lib/openbabel/3.1.0 command=` python ${SurfDockdir}/model/comp_surface/protein_process/openbabel_reduce_openbabel.py \ --data_path ${data_dir} \ --save_path ${surface_out_dir}` state=$command fi #------------------------------------------------------------------------------------------------# #----------------------------- Step2 : Compute Target Surface -----------------------------------# #------------------------------------------------------------------------------------------------# cd $surface_out_dir command=` python ${SurfDockdir}/model/comp_surface/prepare_target/computeTargetMesh_test_samples.py \ --data_dir ${data_dir} \ --out_dir ${surface_out_dir} \ ` state=$command #------------------------------------------------------------------------------------------------# #-------------------------------- Step3 : Get Input CSV File -----------------------------------# #------------------------------------------------------------------------------------------------# command=` python \ ${SurfDockdir}/model/inference_utils/construct_csv_input.py \ --data_dir ${data_dir} \ --surface_out_dir ${surface_out_dir} \ --output_csv_file ${out_csv_file} \ --Screen_ligand_library_file ${Screen_lib_path} \ ` state=$command #------------------------------------------------------------------------------------------------# #-------------------------------- Step4 : Get Pocket ESM Embedding ----------------------------# #------------------------------------------------------------------------------------------------# esm_dir=${SurfDockdir}/model/esm sequence_out_file="${esmbedding_dir}/test_samples.fasta" protein_pocket_csv=${out_csv_file} full_protein_esm_embedding_dir="${esmbedding_dir}/esm_embedding_output" pocket_emb_save_dir="${esmbedding_dir}/esm_embedding_pocket_output" pocket_emb_save_to_single_file="${esmbedding_dir}/esm_embedding_pocket_output_for_train/esm2_3billion_pdbbind_embeddings.pt" # get faste sequence command=`python ${SurfDockdir}/model/datasets/esm_embedding_preparation.py \ --out_file ${sequence_out_file} \ --protein_ligand_csv ${protein_pocket_csv}` state=$command # esm embedding preprateion command=`python ${esm_dir}/scripts/extract.py \ "esm2_t33_650M_UR50D" \ ${sequence_out_file} \ ${full_protein_esm_embedding_dir} \ --repr_layers 33 \ --include "per_tok" \ --truncation_seq_length 4096` state=$command # map pocket esm embedding command=`python ${SurfDockdir}/model/datasets/get_pocket_embedding.py \ --protein_pocket_csv ${protein_pocket_csv} \ --embeddings_dir ${full_protein_esm_embedding_dir} \ --pocket_emb_save_dir ${pocket_emb_save_dir}` state=$command # save pocket esm embedding to single file command=`python ${SurfDockdir}/model/datasets/esm_pocket_embeddings_to_pt.py \ --esm_embeddings_path ${pocket_emb_save_dir} \ --output_path ${pocket_emb_save_to_single_file}` state=$command #------------------------------------------------------------------------------------------------# #------------------------ Step5 : Start Sampling Ligand Confromers ----------------------------# #------------------------------------------------------------------------------------------------# diffusion_model_dir=${model_temp}/weight/docking confidence_model_base_dir=${model_temp}/weight/posepredict protein_embedding=${pocket_emb_save_to_single_file} test_data_csv=${out_csv_file} cd ${SurfDockdir}/scripts/bash_scripts/test_scripts mdn_dist_threshold_test=3.0 version=6 dist_arrays=(3) for i in ${dist_arrays[@]} do mdn_dist_threshold_test=${i} command=`accelerate launch \ --multi_gpu \ --main_process_port ${main_process_port} \ --num_processes ${NUM_GPUS} \ ${SurfDockdir}/scripts/inference_accelerate.py \ --data_csv ${test_data_csv} \ --model_dir ${diffusion_model_dir} \ --ckpt best_ema_inference_epoch_model.pt \ --confidence_model_dir ${confidence_model_base_dir} \ --confidence_ckpt best_model.pt \ --save_docking_result \ --mdn_dist_threshold_test ${mdn_dist_threshold_test} \ --esm_embeddings_path ${protein_embedding} \ --run_name ${confidence_model_base_dir}_test_dist_${mdn_dist_threshold_test} \ --project ${project_name} \ --out_dir ${docking_out_dir} \ --batch_size 400 \ --batch_size_molecule 10 \ --samples_per_complex 40 \ --save_docking_result_number 40 \ --head_index 0 \ --tail_index 10000 \ --inference_mode Screen \ --wandb_dir ${temp}/docking_result/test_workdir` state=$command done #------------------------------------------------------------------------------------------------# #---------------- Step6 : Start Rescoring the Pose For Screening -----------------# #------------------------------------------------------------------------------------------------# echo '---------------- Step4 : Start Rescoring the Pose For Screening -----------------' # project_name='SurfDock_Screen_samples/repeat_zero' # surface_out_dir=${SurfDockdir}/model/data/Screen_sample_dirs/${project_name}/test_samples_8A_surface # data_dir=${SurfDockdir}/model/data/Screen_sample_dirs/test_samples out_csv_file=${out_csv_dir}/score_inplace.csv command=` python \ ${SurfDockdir}/model/inference_utils/construct_csv_input.py \ --data_dir ${data_dir} \ --surface_out_dir ${surface_out_dir} \ --output_csv_file ${out_csv_file} \ --Screen_ligand_library_file ${Screen_lib_path} \ --is_docking_result_dir \ --docking_result_dir ${docking_out_dir} \ ` state=$command confidence_model_base_dir=${model_temp}/weight/screen test_data_csv=${out_csv_file} version=6 dist_arrays=(3) for i in ${dist_arrays[@]} do mdn_dist_threshold_test=${i} echo mdn_dist_threshold_test : ${mdn_dist_threshold_test} command=`accelerate launch \ --multi_gpu \ --main_process_port ${main_process_port} \ --num_processes 1 \ ${SurfDockdir}/scripts/evaluate_score_in_place.py \ --data_csv ${test_data_csv} \ --confidence_model_dir ${confidence_model_base_dir} \ --confidence_ckpt best_model.pt \ --model_version version6 \ --mdn_dist_threshold_test ${mdn_dist_threshold_test} \ --esm_embeddings_path ${protein_embedding} \ --run_name ${project_name}_test_dist_${mdn_dist_threshold_test} \ --project ${project_name} \ --out_dir ${docking_out_dir} \ --batch_size 40 \ --wandb_dir ${temp}/wandb/test_workdir` state=$command done cat << 'EOF' ____ _ _ _ _ _ _ ____ _ _____ ____ _ _ ____ _ _ ____ _ ____ _ _ ____ __ ____ _ ____ _ _ ____ _ / ___| _ _ _ __ / _| _ \ ___ ___| | __ / ___| __ _ _ __ ___ _ __ | (_)_ __ __ _ | _ \ ___ _ __ ___| | \___ \| | | | '__| |_| | | |/ _ \ / __| |/ / \___ \ / _` | '_ ` _ \| '_ \| | | '_ \ / _` | | | | |/ _ \| '_ \ / _ \ | ___) | |_| | | | _| |_| | (_) | (__| < ___) | (_| | | | | | | |_) | | | | | | (_| | | |_| | (_) | | | | __/_| |____/ \__,_|_| |_| |____/ \___/ \___|_|\_\ |____/ \__,_|_| |_| |_| .__/|_|_|_| |_|\__, | |____/ \___/|_| |_|\___(_) |_| |___/ ____ _ _ _ _ _ _ ____ _ _____ ____ _ _ ____ _ _ ____ _ ____ _ _ EOF