菜单
Load when removing batch effects from a multi-cohort bulk RNA-seq dataset using ComBat (R or Python implementation). Skip if there is only one batch, or for single-cell batch integration (use sc-batch-integration), or for spatial multi-slice integration (use spatial-integrate).
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Operate MedSAM2 for promptable segmentation of 3D medical images and medical videos, including CT lesion propagation, MRI volumes, RECIST-guided prompts, efficient CPU-oriented variants, training, and 3D Slicer integration. Use when generating or validating volumetric masks from sparse prompts or propagating masks through image slices or video frames.
Build reproducible healthcare imaging pipelines with Project MONAI for DICOM, NIfTI, pathology, and multidimensional imaging tasks including preprocessing, augmentation, training, sliding-window inference, evaluation, model bundles, labeling, and deployment. Use when implementing medical image classification, segmentation, registration, detection, generative, or foundation-model workflows in PyTorch.
Operate Google TxGemma prediction and chat models for therapeutic property prediction across small molecules, proteins, nucleic acids, diseases, targets, and cell lines. Use when formatting Therapeutics Data Commons tasks, choosing TxGemma model size or variant, running local or Model Garden inference, fine-tuning on private therapeutic data, or evaluating TxGemma in drug-discovery workflows.
Evaluate and operate released Profluent OpenCRISPR gene-editing systems, especially OpenCRISPR-1, for controlled research workflows using its published Cas9-like protein, compatible guide RNA designs, protocols, licensing, specificity testing, and experimental validation. Use when comparing OpenCRISPR-1 with SpCas9, planning nonclinical editing studies, or assessing use in nuclease, nickase, deactivated, base, prime, or epigenome-editing contexts.
Operate CZI TranscriptFormer cross-species generative single-cell models to produce cell embeddings, contextual gene embeddings, likelihoods, zero-shot classifiers, disease-state representations, and regulatory analyses from raw-count AnnData files. Use when selecting TF-Sapiens, TF-Exemplar, or TF-Metazoa, processing in- or out-of-distribution species, or scaling embedding extraction across GPUs.
Operate ByteDance Protenix-v2 for open biomolecular structure prediction of proteins, antibodies, nucleic acids, ligands, and complexes using JSON inputs, MSA and template features, constraints, and inference-time sampling. Use when running Protenix locally or through its server, comparing AlphaFold3-style open models, or building reproducible co-folding evaluations.
| name | bulkrna-batch-correction |
| description | Load when removing batch effects from a multi-cohort bulk RNA-seq dataset using ComBat (R or Python implementation). Skip if there is only one batch, or for single-cell batch integration (use sc-batch-integration), or for spatial multi-slice integration (use spatial-integrate). |
| version | 0.3.0 |
| author | OmicsClaw |
| license | MIT |
| tags | ["bulkrna","batch-correction","ComBat","harmonization","batch-effect"] |
Run when bulkrna-qc PCA or sample-correlation heatmap reveals samples
clustering by batch (cohort, sequencing run, library prep date) rather
than by biology. Applies ComBat — preferring the R sva implementation
when available, falling back to a Python port.
| Input | Format | Required |
|---|---|---|
| Expression matrix | .csv (gene × sample) | yes (or --demo) |
| Batch metadata | --batch-info CSV (sample, batch cols) | yes (or --demo) |
--mode | parametric or non-parametric | default parametric |
| Output | Path | Notes |
|---|---|---|
| Corrected matrix | tables/corrected_expression.csv | same shape, batch effect removed |
| PCA before correction | figures/pca_before_correction.png | per-sample colour-coded by batch |
| PCA after correction | figures/pca_after_correction.png | same layout, post-ComBat |
| Batch assessment | figures/batch_assessment.png | side-by-side comparison + silhouette delta |
| Batch metrics | tables/batch_metrics.csv | silhouette score before/after |
| Report | report.md + result.json | summary keys: n_genes, n_samples, n_batches, batch_names, mode, silhouette_before, silhouette_after |
sva::ComBat first; on import failure, fall back to Python ComBat (bulkrna_batch_correction.py:427 warns "R ComBat not available (...); using Python fallback.").:165 ("Only 1 batch detected; returning data unchanged.") when --batch-info describes a single batch. The R path has no equivalent guard.bulkrna_batch_correction.py:165 returns the input unchanged with a warning when only one batch is detected — but only when the Python ComBat path runs. The R sva::ComBat path at :421 does not have this guard, so a single-batch run on an R-equipped system may proceed with nonsense output. Verify result.json["n_batches"] ≥ 2 before trusting downstream results.:427's silent fallback to the Python port can produce per-gene corrected values that differ at the 3rd decimal from R sva — usually inconsequential for downstream DE but visible in direct value comparisons. The chosen backend is not recorded in the summary dict; only the warning log distinguishes them.condition × batch before running, and consider including condition as a covariate in a more sophisticated tool (limma::removeBatchEffect) if confounded.corrected_expression.csv into bulkrna-de (which expects non-negative integer counts) — use the corrected matrix only for visualisation, clustering, or co-expression analysis.silhouette_before / silhouette_after (in result.json and batch_metrics.csv) measure batch clustering tightness. A drop indicates batch effect has been reduced; absolute values depend on how separable the batches were originally.python omicsclaw.py run bulkrna-batch-correction --demo
python omicsclaw.py run bulkrna-batch-correction \
--input counts.csv --batch-info batches.csv --output results/
python omicsclaw.py run bulkrna-batch-correction \
--input counts.csv --batch-info batches.csv --output results/ \
--mode non-parametric
references/parameters.md — every CLI flag and tuning hintreferences/methodology.md — parametric vs non-parametric ComBat, R↔Python differences, design-confound caveatsreferences/output_contract.md — exact output directory layoutbulkrna-qc (run upstream to spot batch effects), bulkrna-coexpression / bulkrna-survival (downstream — corrected matrix safe for these), bulkrna-de (NOT downstream-safe — DE always wants raw counts), sc-batch-integration (single-cell sibling: Harmony/scVI/etc.)