Find open-source science resources

Curated list of atomistic ML projects for materials science

GlycoTraitR is an R package for analyzing glycoproteomics data, particularly glycopeptide-spectrum matches (GPSMs). It supports results generated by the pGlyco3 and Glyco-Decipher search engines. The package parses glycan structures, computes monosaccharide compositions and structural traits, and performs differential analysis of glycan heterogeneity. It constructs trait-by-PSM matrices stored in a SummarizedExperiment object, supports user-defined structural motifs, and provides visualization utilities for interpreting glycan trait changes.

Biological simulation tools

Differentiable PDE solving framework for machine learning with built-in fluid simulation, supporting PyTorch/JAX/TensorFlow backends and enabling neural network training within physical simulations (TUM, MIT License)

Inflexa is an open-source, agentic orchestration platform for computational biology and translational medicine. It is designed to assist researchers in analyzing multi-omics, cheminformatics, and imaging data by reading published literature, designing multi-step analysis plans, and executing experiments with full reproducibility.

Flow-based generative model for atomistic protein binder design with test-time optimization, SOTA on binder benchmarks (ICLR 2026 Oral, NVIDIA)

Latent-space probabilistic denoising diffusion model for predicting coarse-grained conformational ensembles of intrinsically disordered proteins and regions from sequence, with GPU/CPU inference, trajectory export, and FAISS-based similarity search (67+ stars, LGPL-3.0)

JavaScript genome browser that is highly customizable via plugins and track customizations.

First scientific ML benchmark with paired real-world measurements and matched numerical simulations for complex physical systems, featuring 5 scenarios, 700+ trajectories, 10 baseline models, and 9 evaluation metrics with HuggingFace datasets and model checkpoints (Westlake University, CC BY-NC 4.0)

Provides a graphical user interface to integrate, visualize and explore results from linkage and quantitative trait loci analysis, together with genomic information for autopolyploid species. The app is meant for interactive use and allows users to optionally upload different sources of information, including gene annotation and alignment files, enabling the exploitation and search for candidate genes in a genome browser. In its current version, 'VIEWpoly' supports inputs from 'MAPpoly', 'polymapR', 'diaQTL', 'QTLpoly', 'polyqtlR', 'GWASpoly', and 'HIDECAN' packages.

Large-scale knowledge graph and pip-installable client for literature-grounded automated scientific research, connecting papers, authors, institutions, venues, keywords, citations, and a four-level research taxonomy across medicine, social sciences, engineering, computer science, materials science, and more (ZJU NLP, arXiv 2026, 136+ stars, MIT License)

A molecule manipulation library.

AI co-author covering the entire research lifecycle — from an under-specified research direction to a published paper; includes ResearchStudio-Idea for evidence-grounded research ideation and ResearchStudio-Reel for turning finished papers into posters, narrated videos, blogs, and interactive reels; runs as skills on Claude Code and Codex (1.2K+ stars, MIT License, 2026)

LLMs as copilots for theorem proving in Lean 4, exposing native tactics (`suggest_tactics`, `search_proof`, `select_premises`) that embed language model inference and premise retrieval directly inside the Lean proof environment, supporting local CTranslate2/CUDA inference as well as remote model APIs for interactive and automated proof search (Caltech & NVIDIA, NeurIPS 2024, 1.2K+ stars)

Galaxy workflow for BlockClust pipeline.

Regional Association of Methylome variability with the Exposome and geNome (RAMEN) is an R package whose goal is to identify genome-wide Variable Methylated Loci (VML) from microarray DNA methylation data; then, using genomic and exposomic data, it can identify which model out of the following explains best the DNA methylation variability at each VML: genetic (G), environmental (E), additive (G+E) or interactive (GxE).

LLM-driven machine learning engineering agent using agentic tree search to autonomously draft, debug and benchmark ML code; wins 4× more medals than the best linear agent on OpenAI's MLE-Bench (75 Kaggle competitions) (1.3K+ stars, MIT License)

TADCompare is an R package designed to identify and characterize differential Topologically Associated Domains (TADs) between multiple Hi-C contact matrices. It contains functions for finding differential TADs between two datasets, finding differential TADs over time and identifying consensus TADs across multiple matrices. It takes all of the main types of HiC input and returns simple, comprehensive, easy to analyze results.

Generalized Attribute Based Ratings Information Extraction Library; official OpenAI toolkit that turns messy qualitative corpora into analysis-ready datasets for social scientists and data scientists, measuring quantitative attributes in text, images, or audio using the GPT API. See the [official blog post](https://openai.com/index/scaling-social-science-research/) and [NBER working paper](http://www.nber.org/papers/w34834) (413+ stars, Apache 2.0)

RFdiffusion is an open source method for structure generation, with or without conditional information (a motif, target etc).

Quantify expression of transposable elements (TEs) from RNA-seq data through different methods, including ERVmap, TEtranscripts and Telescope. A common interface is provided to use each of these methods, which consists of building a parameter object, calling the quantification function with this object and getting a SummarizedExperiment object as output container of the quantified expression profiles. The implementation allows one to quantify TEs and gene transcripts in an integrated manner.

Filter genetic variants using different criteria such as inheritance model, amino acid change consequence, minor allele frequencies across human populations, splice site strength, conservation, etc.

Gene Set Variation Analysis (GSVA) is a non-parametric, unsupervised method for estimating variation of gene set enrichment through the samples of a expression data set. GSVA performs a change in coordinate systems, transforming the data from a gene by sample matrix to a gene-set by sample matrix, thereby allowing the evaluation of pathway enrichment for each sample. This new matrix of GSVA enrichment scores facilitates applying standard analytical methods like functional enrichment, survival analysis, clustering, CNV-pathway analysis or cross-tissue pathway analysis, in a pathway-centric manner.

Open-source deep learning toolbox for bioimage analysis providing a unified, configuration-driven framework for 2D/3D semantic segmentation, instance segmentation, classification, denoising, super-resolution, and self-supervised learning; integrates state-of-the-art architectures including U-Net, Vision Transformers, and ConvNeXt, designed for microscopy and biomedical imaging researchers without extensive coding expertise (MIT License, actively maintained)

A compressor of common genomic file formats (BAM, CRAM, FASTQ, VCF etc).

Structural variant discovery by integrated paired-end and split-read analysis.

Curated library of 550+ medical research agent skills spanning evidence insights, protocol design, omics/clinical data analysis, and academic writing; each skill is reviewed through MedSkillAudit and compatible with Claude Code, Codex, Open Code, OpenClaw, and SKILL.md-compatible agents (AIPOCH, 1.2K+ stars, MIT License, 2026)

methylKit is an R package for DNA methylation analysis and annotation from high-throughput bisulfite sequencing. The package is designed to deal with sequencing data from RRBS and its variants, but also target-capture methods and whole genome bisulfite sequencing. It also has functions to analyze base-pair resolution 5hmC data from experimental protocols such as oxBS-Seq and TAB-Seq. Methylation calling can be performed directly from Bismark aligned BAM files.

HiCDOC normalizes intrachromosomal Hi-C matrices, uses unsupervised learning to predict A/B compartments from multiple replicates, and detects significant compartment changes between experiment conditions. It provides a collection of functions assembled into a pipeline to filter and normalize the data, predict the compartments and visualize the results. It accepts several type of data: tabular `.tsv` files, Cooler `.cool` or `.mcool` files, Juicer `.hic` files or HiC-Pro `.matrix` and `.bed` files.

the wavefront alignment algorithm (WFA) which expoit sequence similarity to speed up alignment

Advanced OCR with PP-StructureV3 document parsing, 13% accuracy improvement, supports 80+ languages

Autonomous ML experimentation for biomedical data.

Segment Anything Model for microscopy: interactive and automatic segmentation of light, electron, and fluorescence microscopy images in 2D and 3D, with domain-specific fine-tuning workflows for scientific imaging (1.5K+ stars)

A Workflow Management System geared towards scientific workflows.

The miaViz package implements functions to visualize TreeSummarizedExperiment objects especially in the context of microbiome analysis. Part of the mia family of R/Bioconductor packages.

Visualization functions for spatial transcriptomics data. Includes functions to generate several types of plots, including spot plots, feature (molecule) plots, reduced dimension plots, spot-level quality control (QC) plots, and feature-level QC plots, for datasets from the 10x Genomics Visium and other technological platforms. Datasets are assumed to be in either SpatialExperiment or SingleCellExperiment format.

Google Research's hybrid ML/physics atmospheric model combining learned dynamics with physical constraints, outperforming traditional models on 2-15 day forecasts and 40-year climate simulation, developed with ECMWF (Nature 2024)

Statistical methods for differential discovery analyses in high-dimensional cytometry data (including flow cytometry, mass cytometry or CyTOF, and oligonucleotide-tagged cytometry), based on a combination of high-resolution clustering and empirical Bayes moderated tests adapted from transcriptomics.

Method for scalable identification of spatially variable genes (SVGs) in spatially-resolved transcriptomics data. The method is based on nearest-neighbor Gaussian processes and uses the BRISC algorithm for model fitting and parameter estimation. Allows identification and ranking of SVGs with flexible length scales across a tissue slide or within spatial domains defined by covariates. Scales linearly with the number of spatial locations and can be applied to datasets containing thousands or more spatial locations.

Method for identification of spatial domains and spatially-aware clustering in spatial transcriptomics data. The method generates spatial domains with smooth boundaries by smoothing gene expression profiles across neighboring spatial locations, followed by unsupervised clustering. Spatial domains consisting of consistent mixtures of cell types may then be further investigated by applying cell type compositional analyses or differential analyses.

PathBench-MIL is a comprehensive, flexible benchmarking/AutoML framework for multiple instance learning in histopathology. PathBench-MIL is expected to be deprecated and replaced by PathForge.

MetaboLights is one of the main public repositories for storage of metabolomics experiments, which includes analysis results as well as raw data. The MsBackendMetaboLights package provides functionality to retrieve and represent mass spectrometry (MS) data from MetaboLights. Data files are downloaded and cached locally avoiding repetitive downloads. MS data from metabolomics experiments can thus be directly and seamlessly integrated into R-based analysis workflows with the Spectra and MsBackendMetaboLights package.

Py-HLA-Match is a Python library for standardised, rule-based HLA (Human Leukocyte Antigen) matching in retrospective analyses, method development, benchmarking, and in-silico studies in immunogenetics and related fields.

Python package for simulation-based inference enabling likelihood-free Bayesian parameter estimation from scientific simulators, with flexible interfaces for neural posterior estimation, sequential methods, and MCMC/variational backends (Mackelab, 825+ stars)

Predicts the pKa values of ionizable groups in proteins and protein-ligand complexes based in the 3D structure.

Pretrained time series foundation model for long-horizon forecasting across diverse scientific domains including climate variables, biomedical signals, and physical observations; decoder-only Transformer architecture with strong zero-shot generalization (19.8K+ stars, Apache 2.0, 2024-2025)

GAIn is a platform for annotating genetic variants, genomic positions, and regions with reproducible, declarative pipelines using curated Genomic Resource Repositories.

atomate2 is a library of computational materials science workflows.

The package imports data from HoverNet, and ProvGigaPath pipelines. Pipeline output data are hosted in a self-owned online repository. Package functionality conveniently incorporates pipeline data into existing MultiAssayExperiment instances from curatedTCGAData.

Utility functions for working with CONCH data, listing remote files. One function assigns HoverNet nuclei to ProvGigaPath tiles with a scale factor to align coordinates. Provides internal utility functions for 'imageFeatureTCGA' and most functions are not meant for end users.