BioMaster

Load BioMaster as one integrated skill-driven bioinformatics assistant.

BioMaster conversational bioinformatics assistant; starts workflows only after execution intent or empty activation.

Automated and marker-guided single-cell cell type annotation using CellTypist, marker review, reference transfer, and confidence-aware label curation.

Ligand-receptor communication inference in single-cell or spatial data with sender-receiver summaries and cautious interpretation.

Workflow for paired or integrated single-cell RNA and ATAC analysis with multimodal latent spaces and regulatory interpretation.

Workflow for spatial transcriptomics preprocessing, domain detection, deconvolution, neighborhood analysis, and publication-ready spatial maps.

Pseudotime, lineage branching, and state-transition analysis for single-cell data with coherent embeddings and annotations.

Standard scRNA-seq preprocessing and clustering with Scanpy: QC, filtering, normalization, HVG selection, PCA, neighbors, UMAP, and Leiden clustering, producing an analysis-ready AnnData object.

Router/index skill over 1,676 deduplicated biomedical AI agent skills aggregated from 20 repositories into 15 categories; use it to search the index, locate the best-matching skill, fetch its SKILL.md on demand, and follow it.

End-to-end protein binder design via BindCraft AF2 hallucination with built-in validation; runs on Modal or locally and reports per-design QC metrics.

Map query scRNA-seq data onto a pre-trained reference atlas via scArches surgical transfer learning (scVI/scANVI) to obtain a shared latent embedding and transferred cell type labels without retraining the reference.

Integrate multiple scRNA-seq batches to remove batch effects while preserving biological variation, using Harmony, scVI, Seurat anchors, or fastMNN.

Meta-skill that installs the full bioSkills collection (425 skills across 62 categories) into a BioMaster project's bioskills library.

Detect and remove doublets from scRNA-seq data using Scrublet (Python), DoubletFinder (R), or scDblFinder (R).

Automated cell type annotation for preprocessed single-cell data using reference-based and custom classifiers (CellTypist, SingleR, Azimuth, scPred), with confidence filtering, consensus voting, and marker-based validation.

Infer and quantify cell-cell communication from scRNA-seq data using CellChat, NicheNet, and LIANA frameworks.

Segment cells from IMC images using Cellpose/Mesmer/steinbock and extract per-cell expression data with spatial coordinates.

Integrate multiple histone modification ChIP-seq tracks into chromatin states via ChromHMM (with alternatives Segway, EpiSegMix, IDEAS, EpiLogos, full-stack ChromHMM).

Cluster and phenotype high-dimensional flow/mass cytometry data to discover cell populations without predefined gates.

Single-cell clustering workflow: PCA dimensionality reduction, k-NN neighbor graph, Leiden/Louvain community detection, UMAP/tSNE embedding, and optional PAGA graph abstraction. Covers Scanpy (Python) and Seurat (R).

Infer cis-regulatory peak-peak (and peak-gene) co-accessibility connections from scATAC data using Cicero, ArchR, or SCENIC+, with Hi-C concordance validation.

Build weighted gene co-expression networks (WGCNA) to detect co-regulated gene modules, correlate them with sample traits, and identify hub genes; includes CEMiTool, hdWGCNA (single-cell), and PyWGCNA alternatives.

Analyze combinatorial CRISPR screens (Big Papi paired-Cas9 or in4mer/Inzolia Cas12a multiplex) to score synthetic-lethal and synthetic-rescue genetic interactions between gene pairs.

Build a tissue/condition-specific metabolic model by constraining a generic genome-scale model with transcriptomics data using GIMME, iMAT, or GTEx-based tissue extraction, then validate against the original model.

End-to-end pooled and single-cell CRISPR screen pipeline: library validation, guide counting, six-stage QC, copy-number/batch correction, design-matched hit calling, and tier-based consensus.

Read, write, create, merge, and convert single-cell data objects (AnnData/Scanpy and Seurat) for downstream analysis.

Sequence-based deep learning (chromBPNet, tangermeme, TF-MoDISco) for ATAC-seq: Tn5 bias correction, variant effect prediction, and de novo motif discovery.

Choose and produce publication-quality 2D dimensionality-reduction plots (PCA, t-SNE, UMAP, PHATE) with deliberate hyperparameters and honest interpretation limits.

Identify differentially methylated regions (DMRs) from WGBS or methylation-array data using tiling, smoothing, or kernel-based approaches, then refine, annotate, visualize, and export them.

Detect and remove cell doublets from flow cytometry or CyTOF data using scatter gating, DNA/event-length methods, or regression residuals, with batch processing and visualization.

Predict which gene a distal accessible (enhancer) region regulates by combining accessibility activity, 3D contact frequency, and sequence features into a per-(enhancer, gene) score; validate with CRISPRi-FlowFISH.

Detect and remove cell doublets/aggregates from flow cytometry or CyTOF data using scatter gating, automated/QC methods, regression/ratio scoring, and CyTOF DNA/event-length detection, before clustering or quantitative analysis.

Infer gene regulatory networks from single-cell data (pySCENIC for RNA-only, SCENIC+ for Multiome) and simulate TF perturbations with CellOracle.

Estimate SNP heritability and partition it across functional categories, cell types, and loci using LDSC, LDAK SumHer, HDL, and HESS.

Interactively annotate cell types in multiplexed imaging (IMC) data using napari visualization with marker overlays, then extract training data, propagate labels with KNN, and validate annotations.

Reconstruct cell lineage trees from CRISPR/lentiviral/mitochondrial barcodes and analyze clonal dynamics and fate decisions in single-cell lineage-tracing experiments.

Find differentially expressed marker genes per cluster, visualize them, score gene sets/cell cycle, and manually annotate cell types. Supports Scanpy (Python) and Seurat (R).

Build publication-ready figures in Python with matplotlib's object-oriented Figure/Axes API, seaborn integration, Type-42 fonts, CVD-safe palettes, and rasterized point layers.

Infer metabolite-mediated cell-cell communication from scRNA-seq data using MeboCost, by predicting metabolite secretion from enzyme expression and sensing via receptors.

Compute per-sample/per-cell TF motif accessibility deviation z-scores with chromVAR (bulk, Signac, ArchR) and optionally refine TF activity with DecoupleR.