biomed.sm.mv te 84m Molecule Net ligand_scaffold M U V 101

Introduction

The IBM BIOMED.SM.MV-TE-84M-MOLECULENET-LIGAND_SCAFFOLD-MUV-101 model is a multimodal biomedical foundation model designed for small molecule analysis. It utilizes the MMELON (Multi-view Molecular Embedding with Late Fusion) approach to integrate multiple molecular representations (sequence, image, graph) for robust predictive performance across various tasks, such as ligand-protein binding and molecular property prediction.

Architecture

The model employs a multi-view strategy combining image, graph, and text representations:

• Image Representation: Captures 2D molecular structure using RDKit and data augmentation.
• Graph Representation: Encodes molecules as graphs with atoms as nodes and bonds as edges, using categorical embedding for properties.
• Text Representation: Uses SMILES strings tokenized with a custom tokenizer and embedded through a transformer architecture.

These embeddings are aggregated through an attention-based module to form a unified multi-view embedding, boosting performance on predictive tasks.

Training

The model can be fine-tuned for various tasks, including regression and classification, such as binding affinity and toxicity prediction. It supports using pre-trained embeddings for chemical similarity measures and can be combined with protein embeddings for specific tasks.

Guide: Running Locally

Prerequisites

• Operating System: Linux or macOS
• Python Version: Python 3.11
• Conda: Anaconda or Miniconda installed
• Git: For cloning the repository

Installation Steps

1. Set Up the Project Directory

   export ROOT_DIR=~/biomed-multiview
   mkdir -p $ROOT_DIR
   

2. Create and Activate a Conda Environment

   conda create -y python=3.11 --prefix $ROOT_DIR/envs/biomed-multiview
   conda activate $ROOT_DIR/envs/biomed-multiview
   

3. Clone the Repository

   mkdir -p $ROOT_DIR/code
   cd $ROOT_DIR/code
   git clone https://github.com/BiomedSciAI/biomed-multi-view.git
   cd biomed-multi-view
   

4. Install Package Dependencies

   pip install -e .['dev']
   pip install -r requirements.txt
   

5. macOS-Specific Instructions

   For Apple Silicon, disable globbing and install macOS-specific packages:

   noglob pip install -e .[dev]
   pip install -r requirements-mps.txt
   

6. Installation Verification (Optional)

   python -m unittest bmfm_sm.tests.all_tests
   

Cloud GPUs

For improved performance, consider utilizing cloud GPUs from providers such as AWS, GCP, or Azure.

License

The model is distributed under the Apache 2.0 License. For more details, refer to Apache License 2.0.