Find open-source science resources

SSSOM is a Simple Standard for Sharing Ontological Mappings, providing - a TSV-based representation for ontology term mappings - a comprehensive set of standard metadata elements to describe mappings and - a standard translation between the TSV and the Web Ontology Language (OWL). Most metadata elements, such as "sssom:mapping_justification" are defined in the sssom namespace.

MS-based metabolomics data processing and compound annotation pipeline.

Performs feature annotations on LC-MS All-ion fragmentation datasets using fragment ion libraries.

Decentralized self-organizing teams of AI agents for long-running computational scientific experimentation; agents critique each other's proposals before spending compute and share successes/failures to avoid redundant exploration, achieving +8.33% on BioML-Bench, 1.9× faster nanoGPT optimization, and +12.5% on ProteinGym ACE2-Spike (425+ stars, 2026)

Generative foundation model for functional antibody and nanobody design, supporting de novo generation, affinity maturation, inverse design, structure prediction, and humanization (Tencent AI4S, ICLR 2025)

For a convenient overview and download list, visit our model page for this model.

With the dedicated fortify method implemented for flowSet, ncdfFlowSet and GatingSet classes, both raw and gated flow cytometry data can be plotted directly with ggplot. ggcyto wrapper and some customed layers also make it easy to add gates and population statistics to the plot.

A 53K-parameter weight-shared transformer that learns SMILES grammar by applying one small block 8 times. It reaches 95.3% validity on ZINC-250K — outperforming an unshared GPT 10× larger (87.6%).

Single-cell RNA-sequencing (scRNA-seq) has made it possible to profile gene expression in tissues at high resolution. An important preprocessing step prior to performing downstream analyses is to identify and remove cells with poor or degraded sample quality using quality control (QC) metrics. Two widely used QC metrics to identify a ‘low-quality’ cell are (i) if the cell includes a high proportion of reads that map to mitochondrial DNA encoded genes (mtDNA) and (ii) if a small number of genes are detected. miQC is data-driven QC metric that jointly models both the proportion of reads mapping to mtDNA and the number of detected genes with mixture models in a probabilistic framework to predict the low-quality cells in a given dataset.

Provides univariate and multivariate statistics for feature prioritization in untargeted LC-MS metabolomics research.

This set of model weights was released with the GitHub-compatible esm package format. The models here are kept for backwards compatibility, but we recommend you use the HuggingFace-compatible model weights at biohub/ESMC-6B (or biohub/ESMC-300M / biohub/ESMC-600M) instead.

Fit linear models to overdispersed count data. The package can estimate the overdispersion and fit repeated models for matrix input. It is designed to handle large input datasets as they typically occur in single cell RNA-seq experiments.

The package coalesces typical helper functions that are scattered throughout the Bioconductor ecosystem. It aims to reduce code redundancy by formalizing functions often used by Bioconductor developers. These functions include operations such as replacing slots in an object, selecting observations for show methods, labeling function life cycles, and more.

The package provides a set of functions to interact with the Google Cloud Platform (GCP) services on the AnVIL platform. The package is designed to use the API calls from the AnVIL package. It coordinates AnVIL workspace functionality with native GCP tools.

Lower-level functionality to interface with Google Cloud Platform tools. 'gcloud' and 'gsutil' are both supported. The functionality provided centers around utilities for the AnVIL platform.

The AnVIL is a cloud computing resource developed in part by the National Human Genome Research Institute. The AnVILAz package supports end-users and developers using the AnVIL platform in the Azure cloud. The package provides a programmatic interface to AnVIL resources, including workspaces, notebooks, tables, and workflows. The package also provides utilities for managing resources, including copying files to and from Azure Blob Storage, and creating shared access signatures (SAS) for secure access to Azure resources.

# Geneformer Geneformer is a foundational transformer model pretrained on a large-scale corpus of human single cell transcriptomes to enable context-aware predictions in settings with limited data in network biology.

End-to-end autonomous AI research engine that turns an idea into a complete LaTeX paper by dispatching real computational experiments to local GPUs or SLURM clusters, collecting actual results, generating figures/tables, and writing a data-grounded manuscript rather than LLM hallucinations (OpenRaiser, 1.5K+ stars, MIT License, 2026)

A vocabulary for describing semantic assets, defined as highly reusable metadata (e.g. XML1 schemata, generic data models) and reference data (e.g. code lists, taxonomies, dictionaries, vocabularies).

Provide infrastructure to store and access genomewide position-specific scores within R and Bioconductor.

Curated collection of agent skills for scientific research (InternScience, 493+ stars, 2026)

Sparse Autoencoder (SAE) trained on residue-level embeddings from ESM-2 (650M, layer 33) for interpretability research on protein language models.

Implements supervised cell type-aware non-negative matrix factorization (NMF) for dimensional reduction in single-cell RNA sequencing analysis. The package provides methods for incorporating cell type information into the dimensionality reduction process, enabling improved visualization and downstream analysis of single-cell data while preserving biological structure. CellMentor employs a unique loss function that simultaneously minimizes variation within known cell populations while maximizing distinctions between different cell types, enabling effective transfer of learned patterns from labeled reference datasets to new unlabeled data.

DISCO (DIffusion for Sequence-structure CO-design) is a multimodal generative model that simultaneously co-designs protein sequences and 3D structures, conditioned on and co-folded with arbitrary biomolecules — including small-molecule ligands, DNA, and RNA.

Vendors an assortment of useful header-only C++ libraries. Bioconductor packages can use these libraries in their own C++ code by LinkingTo this package without introducing any additional dependencies. The use of a central repository avoids duplicate vendoring of libraries across multiple R packages, and enables better coordination of version updates across cohorts of interdependent C++ libraries.

Generalized biological foundation model with unified nucleic acid and protein language, integrating DNA/RNA/protein sequences (Nature Machine Intelligence 2025)

These tools facilitate batch effects analysis and correction in high-throughput experiments. It was developed primarily for mass-spectrometry proteomics (DIA/SWATH), but could also be applicable to most omic data with minor adaptations. The package contains functions for diagnostics (proteome/genome-wide and feature-level), correction (normalization and batch effects correction) and quality control. Non-linear fitting based approaches were also included to deal with complex, mass spectrometry-specific signal drifts.

Automates and standardizes ligand preparation for AutoDock Vina.

A PyTorch MLP that predicts odor descriptors from a molecule's SMILES string using Morgan (ECFP4) fingerprints. Given any molecule, the model outputs a smell profile across 50 odor categories.

COMO (Closed-loop Optical Molecule recOgnition) is a deep learning framework for Optical Chemical Structure Recognition (OCSR). It recognizes chemical structure diagrams from images and predicts SMILES strings with atom-level 2D coordinates and bond matrices.

Simulation of large-scale brain models

Modelo fine-tunado com LoRA (MLX / Apple Silicon) para assistência clínica em saúde da mulher.

3D vision-language model for computed tomography that leverages both structured electronic health records (EHR) and unstructured radiology reports for pretraining, enabling multimodal medical understanding and radiology report generation (447+ stars, MIT License, 2026)

Automated cell type annotation tool for single-cell transcriptomics using gradient boosting and logistic regression with reference atlases, enabling standardized classification across datasets (Wellcome Sanger Institute, Nature Biotechnology 2022)

vitreg4so150mp14ls_dino-v2-bio is a Bio-DINO image encoder for natural photographs of living organisms. It uses a SoViT-150M/14 Vision Transformer with 4 register tokens and 133.6M backbone parameters, trained with a DINOv2-style self-supervised objective on approximately 31 million curated images…

vitreg1s14lsdino-v2-dist-bio is a compact Bio-DINO image encoder distilled from the larger Bio-DINO SoViT-150M/14 model. It keeps the same natural-photography biodiversity scope as the teacher model, but uses a much smaller ViT-S/14-style student with 21.7M backbone parameters and 384-dimensional…

Multimodal LLM-based AI agent enabling deep research in spatial transcriptomics, automating analysis and interpretation of spatial gene expression data (Harvard LiuLab, bioRxiv 2025)

Automate downloading, opening, and parsing DrugBank.

LLM agents for working with the SRA (Sequence Read Archive) and associated bioinformatics databases, enabling natural language querying of high-throughput sequencing data and metadata across genomic repositories (Arc Institute, 169+ stars, 2024-2026)

BED files store ranged genomic data that can be queried even when the files are compressed. iscream can query data from BED files and return them in muliple formats: parsed records or their summary statistics as data frames or GenomicRanges objects, and matrices as matrix, GenomicRanges, or SummarizedExperiment objects. iscream also provides specialized support for importing methylation data.

This is the pretrained VetBERT model from the github repo: https://github.com/havocy28/VetBERT

A package that allows interactive exploration of AnnotationHub and ExperimentHub resources. It uses DT / DataTable to display resources for multiple organisms. It provides template code for reproducibility and for downloading resources via the indicated Hub package.

Robust normalization and difference calling procedures for ChIP-seq and alike data. Read counts are modeled jointly as a binomial mixture model with a user-specified number of components. A fitted background estimate accounts for the effect of enrichment in certain regions and, therefore, represents an appropriate null hypothesis. This robust background is used to identify significantly enriched or depleted regions.

This package allows to efficiently obtain count vectors from indexed bam files. It counts the number of reads in given genomic ranges and it computes reads profiles and coverage profiles. It also handles paired-end data.

This repository contains the full Bio-DINO DINOv2 training weights for a SoViT-150M/14 Vision Transformer trained on natural photographs of living organisms. It is the companion release to the Birder backbone checkpoints at .

DeepMind's neural network for ab-initio quantum chemistry, directly solving the many-electron Schrödinger equation via variational Monte Carlo with antisymmetric wavefunctions, extended to excited states (Phys. Rev. Research 2020, Science 2024)

Standalone browser-based Gene Ontology network viewer for exploring, filtering, searching, and exporting GO term and gene annotation neighborhoods from locally preprocessed GO OBO and GAF data.