/tools

The DeepMind Research repository contains implementations and code for various research projects, including those focused on molecular simulations and protein structure prediction. It aims to accelerate scientific progress by providing tools and datasets for the research community.

The 'esm' repository provides pretrained transformer models for proteins, enabling tasks such as protein design and structure prediction. It includes tools for generating protein structures from sequences and predicting the effects of mutations, making it a valuable resource in molecular biology.

The 'Awesome AI for Science' repository is a curated collection of resources related to AI applications in various scientific fields, particularly focusing on drug discovery and molecular design. It includes models, datasets, and frameworks that facilitate the prediction and generation of molecular properties and structures.

RFdiffusion is an open-source method for generating protein structures, allowing for both unconditional and motif-based design. It supports various protein design challenges, including binder design and scaffold generation, making it a valuable tool in molecular biology and computational chemistry.

RoseTTAFold is a deep learning package that provides models and scripts for predicting protein structures and their interactions. It utilizes a three-track neural network architecture to enhance the accuracy of these predictions, making it a valuable tool in molecular biology and computational chemistry.

The ESM repository provides flagship generative models for protein sequences, allowing users to generate and predict protein structures and functions. It includes tools for both protein design and representation learning, making it a valuable resource in molecular biology.

This repository is a curated list of papers that explore the use of deep learning techniques in protein design. It includes resources on benchmarks and datasets relevant to the field, making it a valuable tool for researchers in computational biology.

Protenix is an open-source tool designed for high-accuracy biomolecular structure prediction, particularly for proteins. It includes related projects for protein design and benchmarking, enhancing its utility in computational biology and drug discovery.

Graphein is a protein and interactomic graph library that enables the creation of geometric representations of protein and RNA structures, as well as biological interaction networks. It supports various molecular types and provides functionalities for graph construction, visualization, and analysis, making it a valuable resource for molecular design and drug discovery.

PaddleHelix is a bio-computing platform that leverages deep learning for drug discovery, vaccine design, and precision medicine. It offers various applications including molecular property prediction, drug-target interaction prediction, and molecular generation, along with advanced protein structure prediction capabilities.

BindCraft is a user-friendly pipeline for designing protein binders using advanced techniques like AlphaFold2 backpropagation and MPNN. It allows users to specify targets and generates multiple binder designs for experimental characterization.

ColabDesign is a tool that makes protein design accessible through Google Colab, utilizing models like TrRosetta and AlphaFold for generating and optimizing protein structures based on sequences. It provides various functionalities for predicting protein structures and sequences, making it a valuable resource in the field of molecular biology.

BoltzGen is a tool for designing and generating protein structures based on specified design criteria. It utilizes advanced algorithms to produce ranked sets of protein designs, facilitating the exploration of novel molecular architectures.

BioEmu is a tool that emulates protein equilibrium ensembles by sampling structures based on amino acid sequences using generative deep learning techniques. It includes features for steering samples towards physically plausible conformations and can be used for side-chain reconstruction and molecular dynamics equilibration.

Foundry is a central repository for biomolecular foundation models that provides tools for training and using models for protein design, including generative models and structure prediction. It relies on AtomWorks for manipulating biomolecular structures and supports various protein design tasks.

PyRosetta.notebooks offers Jupyter Notebooks that serve as a learning resource for the PyRosetta platform, which is used for biomolecular structure prediction and design. The repository includes tutorials on protein folding, docking, and design, making it a valuable tool for researchers in computational biology and chemistry.

DL4Proteins provides a series of Jupyter notebooks that teach deep learning methodologies for predicting and designing biomolecular structures, particularly proteins. It covers advanced topics such as AlphaFold and graph neural networks, making it a valuable resource for researchers in the field of protein engineering.

FastFold is a high-performance implementation of the Evoformer model used in AlphaFold, aimed at optimizing protein structure prediction on GPU clusters. It enhances training and inference speed while supporting large protein sequences, making it a valuable tool for researchers in molecular biology and computational chemistry.

SaProt is a protein language model that utilizes a structure-aware vocabulary to predict mutational effects and generate protein embeddings. It is designed to enhance the understanding of protein sequences and structures, making it a valuable tool for researchers in molecular biology.

LigandMPNN is a package that offers inference code for generating and redesigning molecular sequences using machine learning models. It supports various configurations for designing ligands and proteins, making it a valuable tool in molecular design.

AlphaFlow is a modified version of AlphaFold that employs flow matching to generate protein conformational ensembles. It is capable of modeling both experimental and molecular dynamics ensembles, providing tools for protein design and simulation.

The RFDiffusionAA repository provides tools for designing small molecule binders and proteins using a diffusion-based approach. It includes functionalities for generating molecular structures and optimizing their properties, making it a valuable resource in molecular design.

RFantibody is a pipeline for the structure-based design of de novo antibodies and nanobodies. It utilizes methods for protein backbone design, sequence design, and in silico filtering of designs, making it a comprehensive tool for antibody engineering.

RFdiffusion2 is an open-source tool designed for enzyme design and molecular generation using advanced inference techniques. It supports the creation of protein structures from atomic motifs and includes benchmarking capabilities for evaluating design performance.