菜单
Load when removing ambient RNA contamination from droplet-based scRNA-seq using a simple subtraction path, CellBender, or SoupX. Skip when the contamination is multiplet barcodes (use sc-doublet-detection) or before counts exist (use sc-count).
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
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。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/mdbabumiamssm/LLMs-Universal-Life-Science-and-Clinical-Skills- --skill sc-ambient-removal命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
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.
基于 SOC 职业分类
正在显示 SKILL.md
| name | sc-ambient-removal |
| description | Load when removing ambient RNA contamination from droplet-based scRNA-seq using a simple subtraction path, CellBender, or SoupX. Skip when the contamination is multiplet barcodes (use sc-doublet-detection) or before counts exist (use sc-count). |
| version | 0.3.0 |
| author | OmicsClaw |
| license | MIT |
| tags | ["singlecell","scrna","ambient","cellbender","soupx","contamination"] |
| requires | ["anndata","numpy","scipy"] |
The user has filtered (or raw + filtered) droplet-based scRNA-seq counts
and suspects ambient RNA from cell-free droplets is inflating per-cell
expression — typical for 10X data with high droplet density. Three
backends share the CLI: simple (a deterministic ambient-profile
subtraction, default), cellbender (Python, requires GPU for sensible
runtime), and soupx (R via rpy2; needs raw + filtered matrices).
Doublets are a different problem — use sc-doublet-detection for
multiplet barcodes.
| Input | Format | Required |
|---|---|---|
| Filtered single-cell AnnData | .h5ad | yes (unless --demo) |
| Raw matrix (SoupX) | 10X mtx dir via --raw-matrix-dir | only for --method soupx |
| Output | Path | Notes |
|---|---|---|
| Cleaned AnnData | processed.h5ad | adata.X overwritten with corrected counts; layers["counts"] preserves the pre-correction matrix; uns["ambient_correction"] (simple) / uns["soupx"] / uns["cellbender"] carries diagnostics |
| Diagnostic figures | figures/counts_comparison.png, figures/count_distribution.png, figures/barcode_rank.png | always rendered |
| Report | report.md + result.json | always written |
--contamination is in [0, 1) and --expected-cells is positive when set.--method against METHOD_REGISTRY.simple.layers["counts"] and overwrite adata.X with the corrected counts; record the run params in uns["ambient_correction"|"soupx"|"cellbender"].report.md + result.json.simple. sc_ambient.py:207-208 logs "Requested method '%s' is unavailable (...). Falling back to simple subtraction." when CellBender is not installed or SoupX cannot reach R/rpy2. After every non-simple run, confirm result.json["summary"]["method_used"] matches what you passed via --method — the flag is a request, not a guarantee.--contamination is bounded to [0, 1) (left-inclusive). sc_ambient.py:133-134 checks 0 <= float(args.contamination) < 1 and raises ValueError("--contamination must be between 0 and 1 (for example 0.05).") otherwise. 0 is allowed (degenerate no-op); 1 and 5.0 (the common typo for 0.05) both fail loudly.--expected-cells must be a positive integer. sc_ambient.py:136 raises ValueError. Zero or negative values fail loudly here rather than producing a degenerate run.--raw-matrix-dir and --filtered-matrix-dir silently falls back to simple. sc_ambient.py:851-857 logs "SoupX requires --raw-matrix-dir and --filtered-matrix-dir. Falling back to simple subtraction." and continues with the simple path. result.json records the fallback in summary["fallback_reason"]; CellBender uses just the filtered matrix and the simple path uses neither.# Demo (simple subtraction)
python omicsclaw.py run sc-ambient-removal --demo --output /tmp/sc_ambient_demo
# CellBender on a 10X-filtered AnnData
python omicsclaw.py run sc-ambient-removal \
--input filtered.h5ad --output results/ \
--method cellbender --expected-cells 8000 --contamination 0.05
# SoupX with explicit raw + filtered matrices
python omicsclaw.py run sc-ambient-removal \
--input filtered.h5ad --output results/ \
--method soupx \
--raw-matrix-dir cellranger_out/raw_feature_bc_matrix \
--filtered-matrix-dir cellranger_out/filtered_feature_bc_matrix
references/parameters.md — every CLI flag and per-method tuning hintreferences/methodology.md — when each backend wins, ambient profile derivation, R/Python tradeoffsreferences/output_contract.md — layers["counts"] (pre-correction) and .X (corrected) semantics, per-method uns diagnosticssc-doublet-detection (parallel — multiplet barcodes, complementary contamination class), sc-filter (upstream — cell QC), sc-preprocessing (downstream — normalise/HVG/PCA on the cleaned counts)