/tools

Q6 is a set of molecular dynamics tools designed for free energy calculations, including Free Energy Perturbation (FEP), Empirical Valence Bond (EVB), and Linear Interaction Energies (LIE). It is developed for use in computational chemistry to facilitate molecular simulations and modeling.

USPNet is a tool designed to predict signal peptides in protein sequences using a deep protein language model. It provides a benchmark set for evaluation and allows users to process and predict using their own protein data.

Deep H3 Loop Prediction is a tool that utilizes a deep residual network architecture to predict probability distributions of inter-residue distances and angles specifically for CDR H3 loops in antibodies. It is designed to assist in the modeling and prediction of antibody structures based on sequence input.

This repository contains source code for applying graph neural networks to improve molecular property prediction by leveraging both high-fidelity and low-fidelity data. It includes methods for transfer learning and provides access to multi-fidelity datasets for drug discovery and quantum mechanics.

DiffDock-Pocket is a molecular docking model that utilizes diffusion processes to predict ligand poses in protein binding pockets while accommodating flexible side chains. It provides functionalities for training models, running inference, and visualizing docking results.

AlphaFold 3 is a hackable inference pipeline for predicting the 3D structures of proteins from their amino acid sequences. It allows users to run predictions without the need for large databases or Docker, making it accessible for experimentation and research.

MolFrames is a Python module designed for handling chemical structures in Pandas DataFrames, allowing for efficient operations on large molecular datasets. It includes a pipelining workflow for processing molecular data and integrates with RDKit for chemical functionality.

pAPRika is an advanced toolkit designed for setting up, running, and analyzing free energy molecular dynamics simulations. It facilitates the calculation of binding free energies, making it a valuable resource for researchers in computational chemistry and molecular biology.

Matlab_Real_Space is an electronic structure code designed for calculating properties of atoms, molecules, clusters, and nanocrystals. It allows users to input molecular coordinates, visualize charge densities, and perform calculations related to electronic structures.

CheTo is a tool for Chemical Topic Modeling that utilizes topic modeling techniques from text mining to analyze chemical data. It provides Jupyter notebooks and datasets for exploring molecular datasets, making it useful for researchers in cheminformatics.

ImmunoStruct is a multimodal deep learning framework designed to predict the immunogenicity of peptide-MHC complexes. By integrating sequence, structural, and biochemical information, it enhances the prediction accuracy for both infectious disease epitopes and cancer neoepitopes.

Profile-QSAR is a project that develops multitask machine learning models to predict the activity of compounds across numerous biological assays. It includes scripts for data retrieval, model building, and making predictions, utilizing ChEMBL data for training and validation.

GPDL is a deep learning framework designed for generating novel protein backbones based on specified motifs and sequences. It utilizes a protein language model to optimize the design process, making it a valuable tool for molecular design in protein engineering.

Molecular_VAE_Pytorch is a PyTorch implementation of a Variational Autoencoder designed for automatic chemical design. It utilizes a continuous representation of molecules to generate new molecular structures based on the ChEMBL dataset.

CPDB is a tool that allows users to parse PDB files into structured DataFrames, facilitating the analysis of protein structures. It supports various input methods, including direct file access and retrieval via UniProt IDs, making it versatile for bioinformatics applications.

The 'rxn-ebm' repository implements energy-based modeling to enhance the performance of retrosynthesis models by re-ranking reactant predictions. It leverages machine learning to analyze reaction databases and improve synthesis planning for organic chemistry.

This repository provides code for meta-learning initializations aimed at improving molecular property prediction in low-resource settings. It includes methods for training models on molecular data, specifically utilizing graph representations of molecules.

DeepRank-GNN-esm is a tool that utilizes graph neural networks to score protein-protein complexes, incorporating features from protein language models. It provides functionalities for generating embeddings and predicting interaction scores, making it useful for molecular design and analysis.

Finetune-ESM is a tool for scalable finetuning of protein language models, utilizing advanced training techniques to enhance the prediction of protein functions from sequences. It supports distributed training and reproducibility, making it suitable for bioinformatics applications.

mpi4pyscf is a plugin that enables MPI parallelism for the PySCF quantum chemistry software. It allows users to run molecular simulations more efficiently by leveraging parallel computing capabilities.

Maxsmi is a tool designed for data augmentation in molecular property prediction using deep learning techniques. It provides various strategies for augmenting SMILES representations to enhance the accuracy of predictions for molecular properties and activities.

RetroBridge is a Markov bridge model that facilitates retrosynthesis planning by predicting reactants for given product molecules. It utilizes a generative framework to learn distributions in a discrete state space, achieving state-of-the-art results in retrosynthesis tasks.

LucaPCycle is a dual-channel model developed to predict whether a protein sequence has phosphate-solubilizing functionality and to classify it into one of 31 specific functional types. The tool employs large language models tailored for protein sequences, making it a valuable resource in protein analysis and molecular biology.