/tools

Invariant Point Attention is a standalone PyTorch module designed for coordinate refinement in protein structures, particularly utilized within the AlphaFold2 framework. It allows for the processing of molecular representations to enhance the accuracy of protein folding predictions.

FLEXS is an open-source environment for developing and comparing algorithms for biological sequence design. It allows users to explore fitness landscapes and implement exploration algorithms to optimize sequences based on model-guided predictions.

The Protein-LLM-Survey repository provides a comprehensive survey of large language models used for protein sequence modeling, understanding, and generation. It includes various methods and models that facilitate protein design and prediction of protein properties.

DeeplyTough is a tool designed for learning structural comparisons of protein binding sites using deep learning techniques. It provides built-in support for several benchmark datasets and allows users to evaluate custom datasets, making it a valuable resource for drug discovery and protein design.

MaSIF-neosurf is a computational tool designed for surface-based protein design, specifically targeting protein-ligand interactions. It employs deep learning techniques to analyze and generate molecular surfaces, facilitating the design of proteins that can interact with small molecules.

FAPLM is an efficient PyTorch implementation of state-of-the-art protein language models, designed to optimize memory usage and inference time. It supports various protein sequence tasks and can be utilized for protein design and benchmarking against official implementations.

PXDesign is a model suite for designing protein binders using a diffusion generator and confidence models. It allows users to create and evaluate potential binders for specific protein targets, facilitating molecular design in drug discovery.

Chroma-PyTorch is an implementation of a generative model for proteins using denoising diffusion probabilistic models and graph neural networks. It aims to facilitate the design of proteins, including those that can bind to specific targets like the spike protein of the coronavirus.

Uni-Fold-jax is a trainable implementation of a deep protein folding model based on AlphaFold, allowing users to train their own models for predicting protein structures. It provides tools for preparing data, training models, and inferring protein structures from input sequences.

AbLang is a language model tailored for antibodies that generates representations for antibody sequences and restores missing residues. It leverages a large dataset of antibody sequences to improve predictions and design applications in molecular biology.

ChatDrug is a tool for conversational drug editing that utilizes retrieval and domain feedback to assist in the design and modification of small molecules, peptides, and proteins. It includes datasets for training and evaluation, making it a comprehensive resource for drug discovery.

NoLabs is an open-source biolab that facilitates experiments in bioinformatics and drug discovery. It includes features for protein design, molecular simulations, and a workflow engine to manage various bioinformatics tasks.

AntiFold predicts sequences that fit into antibody variable domain structures, allowing for the design and sampling of antibody sequences with high structural agreement to experimental data. It utilizes the ESM-IF1 model and is fine-tuned on existing antibody structures, making it a valuable tool for antibody design.

FLAb is a dataset designed for training and benchmarking AI models in therapeutic antibody design, offering extensive data on properties such as binding affinity and thermostability. It serves as a centralized resource for researchers in protein design, facilitating the development of optimized antibody candidates.

AbMAP is a Protein Language Model customized for antibodies, designed to predict structure and functional properties while analyzing B-cell repertoires. It utilizes in-silico mutagenesis and provides embeddings for antibody sequences, making it a useful tool in antibody research and design.

The CHEAP repository provides a framework for the joint embedding of protein sequences and structures using compressed representations. It allows users to obtain embeddings that can be utilized in various applications related to protein design and molecular simulations.

SaprotHub is a platform that democratizes protein language model training and sharing, enabling biologists to fine-tune and utilize protein models without requiring advanced machine learning expertise. It supports various protein prediction tasks, facilitating collaboration and resource sharing within the protein research community.

ProSST is an advanced hybrid language model designed for directed protein evolution, enabling zero-shot prediction of mutant effects. It utilizes a pre-trained transformer model to analyze protein sequences and structures, making it a valuable tool for protein design and benchmarking in computational biology.

This repository provides an implementation of a denoising diffusion probabilistic model tailored for the conditional generation of protein distograms. It utilizes advanced generative modeling techniques to potentially enhance protein structure prediction and design.

The Flex ddG Tutorial provides a framework for modeling and predicting changes in binding free energies of proteins upon mutation using the Rosetta software. It includes example scripts for running the protocol and analyzing results, making it a valuable resource for researchers in computational biology and molecular design.

ScanNet is an interpretable geometric deep learning model that predicts binding sites on proteins. It identifies functional sites such as protein-protein and protein-antibody interactions, leveraging the three-dimensional structure of proteins to enhance predictions.

Pallatom is a protein generation model that produces protein structures with all-atom coordinates by modeling the joint distribution of structure and sequence. It employs a novel network architecture to enhance designability, diversity, and novelty in protein design, making it a valuable tool for researchers in molecular biology.

MSA Pairformer is a tool that extracts evolutionary signals from aligned homologous sequences to predict residue-residue interactions in proteins. It provides functionalities for embedding protein sequences and predicting contacts, making it useful for applications in protein design and analysis.

ProteinNPT is a semi-supervised model designed to predict and generate protein properties. It utilizes non-parametric transformers to learn representations of protein sequences and their associated properties, enabling both property prediction and iterative redesign of proteins.