jidushanbojue / Protac-invent

A deep learning framework for accurate reaction prediction and its application on high-throughput experimentation data

PROTAC-invent

Source code for our paper "3D Based Generative PROTAC Linker Design with Reinforcement Learning". The code was built based on Reinvent (https://github.com/MolecularAI/Reinvent), DockStream (https://github.com/MolecularAI/DockStream). Thanks a lot for their sharing.

Figure 1. Illustration of PROTAC-invent model

Figure 2. The general workflow of PROTAC-invent

Figure 3. The workflow of generating and scoring 3D binding conformation of PROTAC

Quick start

Installation

1. Install Conda

2. Clone this Git repository

3. Open a shell, and go to the repository and create the Conda environment:

    $ conda env create -f reinvent.yml
   

4. Activate the environment:

    $ conda activate reinvent.v3.2
   

5. Install in-house reinvent_scoring

    $ cd reinvent_scoring
   
    $ pip install reinvent_scoring-0.0.73_bq-py3-none-any.whl
   

6. Open another shell, and clone in-house DockStream repository

   This is the docking component special for Protac-invent.

    $ conda env create -f environment.yml
   

Usage

1. Edit template Json file (for example in result/LINK_invent/BTK/template.json).

   Templates can be manually edited before using. The only thing that needs modification for a standard run are the file and folder paths. Most running modes produce logs that can be monitored by tensorboard

2. python input.py template.json

Analyse the results

1. tensorboard --logdir "progress.log"

   progress.log is the "logging_path" in template.json

We selected top200 solutions for some specific Ternary Complex, such as BTK(PDB code: 6W8I), BAF(PDB code: 6HAX), BRD4(PDB code: 5T35), the result as follow:

1. For BTK Ternary Complex:

   BTK_poses_top200.sdf: (https://github.com/jidushanbojue/Protac-invent/tree/master/result/LINK_invent/BTK/BTK_poses_top200.sdf)

   BTK_scored_smiles_top200.xlsx (https://github.com/jidushanbojue/Protac-invent/tree/master/result/LINK_invent/BTK/BTK_scored_smiles_top200.xlsx)

2. For BAF Ternary Complex

   BAF_poses_top200.sdf: (https://github.com/jidushanbojue/Protac-invent/tree/master/result/LINK_invent/BAF/BAF_poses_top200.sdf)

   BAF_scored_smiles_top200.xlsx (https://github.com/jidushanbojue/Protac-invent/tree/master/result/LINK_invent/BAF/BAF_scored_smiles_top200.xlsx)

3. For BRD4 Ternary Complex

   BRD4_poses_top200.sdf: (https://github.com/jidushanbojue/Protac-invent/tree/master/result/LINK_invent/5T35/BRD4_poses_top200.sdf)

   BTK_scored_smiles_top200.xlsx (https://github.com/jidushanbojue/Protac-invent/tree/master/result/LINK_invent/5T35/BRD4_scored_smiles_top200.xlsx)

We manually selected 25 solutions for BTK, perform MD simulation and MM-GBSA calculation, the result as follow

26_PROTAC_MM-GBSA_value.xlsx (https://github.com/jidushanbojue/Protac-invent/tree/master/result/LINK_invent/BTK/26_PROTAC_MM-GBSA_value.xlsx)