compbio

Plugin compbio from marketplace ai-skills in CHENyiru3/AI-Skills-Collections. Commands read from the repo's manifest — 3 days ago.

Skills (38)

• ensembl-database

  Query Ensembl REST for genes/transcripts/proteins, sequences, variants/VEP, homology, regions or assembly mappings, preserving species, release, coordinate/strand, identifier-version and evidence contracts.

• geo-database

  Search, select, download, or curate NCBI GEO series/sample/platform data and metadata, including SOFT/series matrices/supplementary files, probe alignment, and study-design provenance.

• human-cell-atlas

  Find, select, and retrieve Human Cell Atlas projects, matrices, files, and associated metadata from the HCA Data Portal and its documented APIs. Use for HCA-specific data access and provenance.

• kegg-database

  Retrieve KEGG entries, pathways, links, sequences, maps, or identifier conversions with the documented REST API, respecting provider access terms, request rate, and operation-specific limits.

• reactome-database

  Retrieve Reactome pathways/entities or submit and interpret Reactome pathway enrichment/expression analyses, with identifier coverage, species, background, FDR, and database-version provenance.

• computational-biology-project-manager

  Initialize, resume, organize, or hand off a computational biology project using an index and current-state recap. Use for project continuity or explicitly requested local skill distillation; not method selection or result interpretation.

• computational-biology-retrospective

  Review computational biology algorithm-development projects and consolidate past experiments into an evidence-linked Markdown experience report. Use for experiment retrospectives, learning from biological or technical failures, and recovering reusable lessons from mixed project folders. Start from t

• critique-bio-manuscript

  Critique a bioinformatics/computational biology manuscript for supported claims, study design, leakage and evaluation limits. Use for an explicit manuscript review; not a rewrite or an assumption that every study needs new biological experiments.

• draft-spatial-methods

  Draft a Methods section for a supplied spatial transcriptomics or single-cell workflow from its code, configuration and notes. Use for evidence-based Methods writing; not running new experiments or assuming a muscle-aging/3D model.

• explain-bio-dl-model

  Explain a supplied deep-learning architecture in biological manuscript prose with accurate data flow and evidence boundaries. Use for model-to-text translation; not inventing biological rationale, validation or causal conclusions.

• scribble

  Investigate or maintain a user-identified Scribble bioinformatics package after its authoritative repository and API are supplied or found. This entry has an unverified tool identity; do not use it to select a generic multi-omics integration method.

• metabolomics-workbench-database

  Find and retrieve Metabolomics Workbench studies, analysis metadata, measurements, RefMet names, or candidate compounds through its documented API. Use for source-backed metabolomics data access and annotation.

• uniprot-database

  Query UniProt entries, sequences, annotations, searches, or cross-database ID mapping through the REST API, including pagination, isoforms, evidence, and mapping-job status.

• archr

  Build or debug an ArchR scATAC-seq workflow: Arrow files, project QC, iterative LSI, clusters, reproducible peaks, motif analysis, or RNA reference integration.

• signac

  Implement or debug Signac ChromatinAssay workflows for scATAC-seq QC, TF-IDF/LSI, accessibility, gene activity, fragments, genomic annotations, and Seurat multimodal integration.

• pydeseq2

  Design, run, or review PyDESeq2 differential-expression analysis on bulk RNA-seq or biologically replicated pseudobulk counts, with sample alignment, estimable designs, explicit contrasts, and FDR interpretation.

• anndata

  Create, read/write, subset, align, concatenate, or validate AnnData/H5AD/Zarr data structures, including sparse/backed storage, views, layers, graphs, and metadata. Use for data-container contracts rather than a general analysis pipeline.

• anndatar

  Read/write H5AD or convert AnnData, Seurat, and SingleCellExperiment objects using the R package anndataR. Use for anndataR-specific interoperability with explicit assay/layer and identifier mapping.

• monocle3

  Implement or review Monocle 3 cell_data_set preprocessing, graph learning, partition-aware rooting/pseudotime, trajectory-associated genes, and model-based expression tests.

• pertpy

  Select, implement, interpret, or debug a pertpy workflow for single-cell perturbation data. Use for a concrete pertpy analysis; not generic Scanpy preprocessing or unrelated prediction frameworks.

• scanpy

  Implement or review Scanpy scRNA-seq QC, preprocessing, feature selection, graphs, clustering, markers, annotation, and visualization with explicit AnnData layer and biological-replicate contracts.

• schard

  Read or convert H5AD data with cellgeni/schard into Seurat or SingleCellExperiment, including its supported spatial conversion helpers. Use for schard-specific interoperability and mapping issues.

• scvi-tools

  Select, configure, train, or debug a specific scvi-tools probabilistic model for single-cell RNA, ATAC, multimodal, spatial, or specialized data, including model-specific data registration and uncertainty.

• seurat

  Implement or review Seurat R single-cell RNA workflows for QC, assays/layers, normalization, PCA/graphs, clustering, markers, annotation, and supported integration or conversion.

• bbknn

  Build or evaluate a batch-balanced neighbor graph with BBKNN in a Scanpy/AnnData workflow, including batch labels, per-batch neighbors, representation choice, and biological-preservation checks.

• harmony

  Configure, tune, or assess Harmony correction of single-cell low-dimensional representations, including supported Seurat or Python integrations and preservation of biological structure.

• scanorama

  Integrate or correct single-cell datasets with Scanorama, handling its matrix/AnnData APIs, gene alignment, corrected embeddings, and biological-preservation checks.

• seurat-v5

  Work with a Seurat v5 project using Assay5 layers, sample integration, WNN multimodal graphs, or sketch/large-data workflows. Use when these v5 contracts are relevant, rather than for every R single-cell task.

• cellxgene-census

  Query a versioned CZ CELLxGENE Census release for cell/gene metadata or expression, including bounded AnnData/SOMA access, dataset feature presence, sparse statistics, and study-aware evaluation splits.

• cellxgene

  Prepare and inspect a local H5AD dataset for the CELLxGENE Annotate viewer, launch/debug the installed viewer, and manage its annotations or sharing configuration. For Census queries use cellxgene-census.

• giotto

  Import, analyze, or visualize spatial genomics data in an identified Giotto/Giotto Suite R workflow, including spatial networks, statistics, and multimodal objects.

• spatialdata

  Read, construct, validate, transform, or analyze SpatialData containers and Zarr stores, including images, labels, points, shapes, and linked AnnData tables. Use for actual SpatialData workflows.

• squidpy

  Analyze spatial neighborhoods, co-occurrence, autocorrelation, images, or spatial plots with Squidpy, including coordinate, library, graph, and permutation contracts.

• stereo-seq

  Analyze or convert Stereo-seq GEM/GEF, square-bin, or cell-bin spatial transcriptomics data while preserving bin/cell identity, coordinates, and count semantics.

• vitessce

  Prepare data and build or debug Vitessce view configurations, widgets, and linked spatial/single-cell visualizations. Use when Vitessce is requested or already part of the project.

• chimerax-mcp

  Install, configure, troubleshoot, and use dovas-net/chimeraX-mcp through MCP or CLI/API access. Use for ChimeraX molecular visualization, hydrogen bonds, salt bridges, van der Waals contacts, electrostatic surfaces, sessions, and requested figure export.

• nextflow

  Create, review, or debug Nextflow DSL2 workflows, process/channel contracts, sample sheets, resources, containers, execution profiles, and resume behavior, including nf-core workflows.

• snakemake

  Create, inspect, or debug Snakemake rules, dependency graphs, configuration, environments, resources, and execution profiles for a reproducible computational workflow.