Find open-source science resources

sBEE is a novel metric designed to robustly evaluate batch mixing in single-cell RNA sequencing (scRNA-seq) data. It assesses integration quality by combining local neighborhood composition and cross-batch distance relationships to provide stable benchmarking across diverse data characteristics.

scUmaper is an R/Seurat-native workflow designed for single-cell transcriptomics analysis, providing automated quality control, biologically grounded doublet removal, and lineage-marker-based cell-type annotation.

MitoChontrol is a cell-type-aware probabilistic framework for single-cell RNA sequencing (scRNA-seq) quality control. It models mitochondrial transcript content using a Gaussian mixture distribution to selectively filter compromised cells while retaining viable populations.

SEAGALL is a deep learning pipeline that uses geometry-regularised autoencoders (GRAE) and explainable graph attention networks (X-GAT) to analyze single-cell data. It quantifies molecular features to explain cell identity and extract stable signatures from multiomics experiments.

spVIPES is a probabilistic framework used for single-cell RNA-sequencing batch correction, designed to decompose unpaired datasets with nonmatching features into shared (conserved) and private (dataset-specific) cellular representations.

causarray is a causal inference framework designed for single-cell genomic data, allowing researchers to robustly uncover causal relationships and treatment effects in complex, high-dimensional biological datasets under unmeasured confounding.

DestinyNet is a multi-task deep-learning framework designed for cell-fate analysis from lineage-tracing single-cell RNA sequencing (LT-scSeq) data. It performs fate clustering, dynamic fate flow modeling, and early-stage fate prediction.