Find open-source science resources

IsoBayes is a Bayesian method to perform inference on single protein isoforms. Our approach infers the presence/absence of protein isoforms, and also estimates their abundance; additionally, it provides a measure of the uncertainty of these estimates, via: i) the posterior probability that a protein isoform is present in the sample; ii) a posterior credible interval of its abundance. IsoBayes inputs liquid cromatography mass spectrometry (MS) data, and can work with both PSM counts, and intensities. When available, trascript isoform abundances (i.e., TPMs) are also incorporated: TPMs are used to formulate an informative prior for the respective protein isoform relative abundance. We further identify isoforms where the relative abundance of proteins and transcripts significantly differ. We use a two-layer latent variable approach to model two sources of uncertainty typical of MS data: i) peptides may be erroneously detected (even when absent); ii) many peptides are compatible with multiple protein isoforms. In the first layer, we sample the presence/absence of each peptide based on its estimated probability of being mistakenly detected, also known as PEP (i.e., posterior error probability). In the second layer, for peptides that were estimated as being present, we allocate their abundance across the protein isoforms they map to. These two steps allow us to recover the presence and abundance of each protein isoform.

distinct is a statistical method to perform differential testing between two or more groups of distributions; differential testing is performed via hierarchical non-parametric permutation tests on the cumulative distribution functions (cdfs) of each sample. While most methods for differential expression target differences in the mean abundance between conditions, distinct, by comparing full cdfs, identifies, both, differential patterns involving changes in the mean, as well as more subtle variations that do not involve the mean (e.g., unimodal vs. bi-modal distributions with the same mean). distinct is a general and flexible tool: due to its fully non-parametric nature, which makes no assumptions on how the data was generated, it can be applied to a variety of datasets. It is particularly suitable to perform differential state analyses on single cell data (i.e., differential analyses within sub-populations of cells), such as single cell RNA sequencing (scRNA-seq) and high-dimensional flow or mass cytometry (HDCyto) data. To use distinct one needs data from two or more groups of samples (i.e., experimental conditions), with at least 2 samples (i.e., biological replicates) per group.

CopyNumberPlots have a set of functions extending karyoploteRs functionality to create beautiful, customizable and flexible plots of copy-number related data.

Test for univariate and bivariate spatial patterns in spatial omics data with single-molecule resolution. The tests implemented allow for analysis of nested designs and are automatically calibrated to different biological specimens. Tests for aggregation, colocalization, gradients and vicinity to cell edge or centroid are provided.

Physics-Informed Neural networks for Advanced modeling in PyTorch

SMBGC Annotation using Neural Networks Trained on Interpro Signatures

Generate QC reports summarizing the output from an alevin, alevin-fry, or simpleaf run. Reports can be generated as html or pdf files, or as shiny applications.

Benchmark evaluating AI agents on complex real-world scientific workflows in terminal environments across life, physical, earth, and mathematical sciences; featured on model cards for Claude Opus 4.7, GPT-5.5, and Gemini 3.1 Pro (200+ stars, Apache 2.0)

Composable computational-science methodology skills for AI research agents emphasizing pre-registration, reproducible workspaces, and red-team review to guard against p-hacking and HARKing; zero third-party dependencies and runs with any agent harness plus a POSIX shell (281+ stars, MIT License, 2026)

Whole-slide pathology foundation model trained on 1.3 billion image tiles from 171K slides using a LongNet-based architecture to encode gigapixel-scale WSIs for cancer subtyping and biomarker prediction (Microsoft Research & Providence, 601+ stars)

Fast sample-swap and relatedness checks on BAMs/CRAMs/VCFs/GVCFs.

A Simulation Tool for Fractured and Deformable Porous Media.

Language agent gymnasium for challenging scientific tasks including DNA manipulation, literature search, and protein engineering

Studies including both microbiome and metabolomics data are becoming more common. Often, it would be helpful to integrate both datasets in order to see if they corroborate each others patterns. All vs all association is imprecise and likely to yield spurious associations. This package takes a knowledge-based approach to constrain association search space, only considering metabolite-function pairs that have been recorded in a pathway database. This package also provides a framework to assess differential association.

High-performance symbolic regression for discovering interpretable scientific equations from data, multi-population evolutionary search with Python/Julia backend, widely used in physics and astronomy (Cambridge, NeurIPS 2023)

pathlinkR is an R package designed to facilitate analysis of RNA-Seq results. Specifically, our aim with pathlinkR was to provide a number of tools which take a list of DE genes and perform different analyses on them, aiding with the interpretation of results. Functions are included to perform pathway enrichment, with muliplte databases supported, and tools for visualizing these results. Genes can also be used to create and plot protein-protein interaction networks, all from inside of R.

The scRNAseqApp is a Shiny app package designed for interactive visualization of single-cell data. It is an enhanced version derived from the ShinyCell, repackaged to accommodate multiple datasets. The app enables users to visualize data containing various types of information simultaneously, facilitating comprehensive analysis. Additionally, it includes a user management system to regulate database accessibility for different users.

The main function is doppelgangR(), which takes as minimal input a list of ExpressionSet object, and searches all list pairs for duplicated samples. The search is based on the genomic data (exprs(eset)), phenotype/clinical data (pData(eset)), and "smoking guns" - supposedly unique identifiers found in pData(eset).

scTypeEval provides tools to evaluate and validate cell type classifications in single-cell transcriptomics when ground truth labels are limited or unavailable. Results are organized in an S4 object that integrates processed data, dimensional reductions, dissimilarity assays, and consistency metrics computed across samples. The workflow includes preprocessing and feature selection, principal component analysis, computation of dissimilarity matrices, internal validation metrics (for example, silhouette-based summaries), and visualization utilities to inspect heatmaps and PCA plots. Functions support common single-cell containers and enable comparison of clustering and labeling strategies across datasets.

A quality control tool for high throughput sequence data.

NVIDIA and King's College London's open-source AI toolkit for healthcare imaging, providing foundational frameworks for medical image annotation (MONAI Label), training (MONAI Core), and deployment (MONAI Deploy) across radiology, pathology, and endoscopy (8K+ stars, Apache 2.0)

Curated, multilingual library of 182 installable AI agent skills for end-to-end academic research spanning literature discovery, scientific writing, grant development, bioinformatics, drug discovery, clinical research, machine learning, and data analysis (779+ stars, MIT License, 2026)

'tidySingleCellExperiment' is an adapter that abstracts the 'SingleCellExperiment' container in the form of a 'tibble'. This allows *tidy* data manipulation, nesting, and plotting. For example, a 'tidySingleCellExperiment' is directly compatible with functions from 'tidyverse' packages `dplyr` and `tidyr`, as well as plotting with `ggplot2` and `plotly`. In addition, the package provides various utility functions specific to single-cell omics data analysis (e.g., aggregation of cell-level data to pseudobulks).

Transcript quantification import from Salmon and other quantifiers with automatic attachment of transcript ranges and release information, and other associated metadata. De novo transcriptomes can be linked to the appropriate sources with linkedTxomes and shared for computational reproducibility.

Estimate variance-mean dependence in count data from high-throughput sequencing assays and test for differential expression based on a model using the negative binomial distribution.

JCVI is a versatile toolkit for comparative genomics analysis. It is a collection of Python libraries to parse bioinformatics files, or perform computation related to assembly, annotation, and comparative genomics.

Principle-first scientific idea discovery framework that extracts reusable principles from public literature and private research materials, composes them into traceable Idea Cards with prior-art comparisons, and exports validation-ready research packs; emphasizes inspectable scientific objects, risk disclosure, and falsification paths (ICML 2026, 411+ stars, MIT License)

High-performance ML research

A tabular style data object where most data is stored outside main memory. A buffer is used to speed up access to data.

A package that extends and improves the functionality of the base affy package. Routines that make heavy use of compiled code for speed. Central focus is on implementation of methods for fitting probe-level models and tools using these models. PLM based quality assessment tools.

Routines for parsing Affymetrix data files based upon file format information. Primary focus is on accessing the CEL and CDF file formats.

A library of core preprocessing routines.

Python wrapper for [samtools](https://github.com/samtools/samtools).

Molecular dynamics in JAX

Foundation AutoResearch Operating System: blueprint-driven runtime for orchestrating AI research workflows from idea generation and experiments to paper writing and peer review (OpenNSWM-Lab, 2.4K+ stars, 2026)

Coordinate-based genomic visualization package for R. It grants users the ability to programmatically produce complex, multi-paneled figures. Tailored for genomics, plotgardener allows users to visualize large complex genomic datasets and provides exquisite control over how plots are placed and arranged on a page.

Comprehensive Claude Code skill suite covering the full academic pipeline from deep research and paper writing to multi-perspective peer review, revision, and finalization; features multi-agent teams, PRISMA systematic review, style calibration, claim-level citation audits, integrity gates, and human-in-the-loop safeguards (38K+ stars, CC BY-NC 4.0, 2026)

Official Jupyter extension with `%%ai` magic commands and sidebar chat assistant, connecting multiple model providers and local inference

102 executable tasks from 44 peer-reviewed papers across 4 disciplines with containerized evaluation

Scikit-learn compatible tabular foundation model for zero-shot classification and regression on mixed-type tabular datasets via in-context learning; applicable to diverse scientific datasets (1.8K+ stars, Apache 2.0)

Provides a high-level R interface to CoreArray Genomic Data Structure (GDS) data files. GDS is portable across platforms with hierarchical structure to store multiple scalable array-oriented data sets with metadata information. It is suited for large-scale datasets, especially for data which are much larger than the available random-access memory. The gdsfmt package offers the efficient operations specifically designed for integers of less than 8 bits, since a diploid genotype, like single-nucleotide polymorphism (SNP), usually occupies fewer bits than a byte. Data compression and decompression are available with relatively efficient random access. It is also allowed to read a GDS file in parallel with multiple R processes supported by the package parallel.

Diffusion-based generative model for inorganic materials design, steering generation by chemistry, symmetry, bulk modulus, band gap, or magnetic properties, 2× more likely to produce stable novel structures than prior methods, experimentally validated with synthesized TaCr₂O₆ (Microsoft, Nature 2025)

Assay for Transpose-Accessible Chromatin using sequencing (ATAC-seq) is a technique to assess genome-wide chromatin accessibility by probing open chromatin with hyperactive mutant Tn5 Transposase that inserts sequencing adapters into open regions of the genome. ATACseqTFEA is an improvement of the current computational method that detects differential activity of transcription factors (TFs). ATACseqTFEA not only uses the difference of open region information, but also (or emphasizes) the difference of TFs footprints (cutting sites or insertion sites). ATACseqTFEA provides an easy, rigorous way to broadly assess TF activity changes between two conditions.

Bioconductor has a rich ecosystem of metadata around packages, usage, and build status. This package is a simple collection of functions to access that metadata from R. The goal is to expose metadata for data mining and value-added functionality such as package searching, text mining, and analytics on packages.

High-accuracy RAG for scientific PDFs with citation support, agentic RAG, and contradiction detection

A python-based workflow manager.

Reinforcement-learning-trained AI agent for treatment reasoning over a universe of 212 biomedical tools, performing multi-step evidence gathering and spawning parallel reasoning branches to reach evidence-grounded clinical decisions (55+ stars, MIT License, 2026)

PinPath enables flexible visualization of (omics) data onto pathways diagrams, allowing users to pinpoint where the relevant changes occur. It supports pathway diagrams from WikiPathways and KEGG, as well as custom GPML and KGML files. Data can be displayed on both native pathway layouts and network representations

Open-source scientific multimodal foundation model built on a 235B MoE LLM and 6B vision encoder, continually pretrained on 5T tokens including 2.5T scientific-domain tokens, with strong results across chemistry, materials, life science, and earth science benchmarks (2025)

This package provides functionality to combine the existing pieces of the transcriptome data and results, making it easier to generate insightful observations and hypothesis. Its usage is made easy with a Shiny application, combining the benefits of interactivity and reproducibility e.g. by capturing the features and gene sets of interest highlighted during the live session, and creating an HTML report as an artifact where text, code, and output coexist. Using the GeneTonicList as a standardized container for all the required components, it is possible to simplify the generation of multiple visualizations and summaries.