菜单
Cell-cell communication analysis for annotated scRNA-seq data using a built-in ligand-receptor scorer, LIANA, CellPhoneDB, CellChat, or a NicheNet R path.
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-cell-communication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。
正在显示 SKILL.md
| name | sc-cell-communication |
| description | Cell-cell communication analysis for annotated scRNA-seq data using a built-in ligand-receptor scorer, LIANA, CellPhoneDB, CellChat, or a NicheNet R path. |
| version | 0.3.0 |
| author | OmicsClaw Team |
| license | MIT |
| tags | ["singlecell","communication","ligand-receptor","liana","cellphonedb","cellchat","nichenet"] |
| metadata | {"omicsclaw":{"domain":"singlecell","allowed_extra_flags":["--cell-type-key","--cellchat-min-cells","--cellchat-prob-type","--cellphonedb-counts-data","--cellphonedb-iterations","--cellphonedb-pvalue","--cellphonedb-threshold","--cellphonedb-threads","--condition-key","--condition-oi","--condition-ref","--method","--nichenet-lfc-cutoff","--nichenet-expression-pct","--nichenet-top-ligands","--receiver","--senders","--species","--r-enhanced"],"param_hints":{"builtin":"[Truncated]","liana":"[Truncated]","cellphonedb":"[Truncated]","cellchat_r":"[Truncated]","nichenet_r":"[Truncated]"},"saves_h5ad":true,"requires_preprocessed":true,"requires":{"bins":"[Truncated]","env":"[Truncated]","config":"[Truncated]"},"emoji":"S","homepage":"https://github.com/OmicsClaw/OmicsClaw","os":["macos","linux"],"install":["[Truncated]"],"trigger_keywords":["cell communication","cell-cell communication","ligand receptor","cellchat","liana","cellphonedb","nichenet"]}} |
cell_type_key-centric interface across all backends.processed.h5ad, report, structured result JSON, README, tables/, figure_data/, and reproducibility bundle.Implemented entry paths:
builtinlianacellphonedbcellchat_rnichenet_rThe wrapper assumes you already have biologically meaningful labels in obs.
| Format | Extension / form | Current wrapper support | Notes |
|---|---|---|---|
| AnnData | .h5ad | yes | current direct input path |
| Demo | --demo | yes | bundled annotated-example fallback |
cell_type, leiden, or another supplied --cell-type-key.figure_data/, gallery plots, processed.h5ad, report.md, and result.json.python omicsclaw.py run sc-cell-communication \
--input data.h5ad --cell-type-key cell_type --output out/
python omicsclaw.py run sc-cell-communication \
--method liana --input data.h5ad --cell-type-key cell_type --output out/
python omicsclaw.py run sc-cell-communication \
--method cellphonedb --input data.h5ad --cell-type-key cell_type \
--cellphonedb-threshold 0.1 --cellphonedb-iterations 1000 --output out/
python omicsclaw.py run sc-cell-communication \
--method cellchat_r --input data.h5ad --cell-type-key cell_type --output out/
python omicsclaw.py run sc-cell-communication \
--method nichenet_r --input data.h5ad --cell-type-key cell_type \
--condition-key condition --condition-oi stim --condition-ref ctrl \
--receiver Monocyte --senders T_cell,B_cell --nichenet-lfc-cutoff 0.25 --output out/
| Parameter | Role | Notes |
|---|---|---|
--method | communication backend | builtin, liana, cellphonedb, cellchat_r, or nichenet_r |
--cell-type-key | grouping column | core public control across all methods |
--species | ligand-receptor resource selector | affects database interpretation and matching |
--cellphonedb-* | CellPhoneDB controls | used only by the cellphonedb path |
--cellchat-prob-type, --cellchat-min-cells | CellChat controls | used only by the cellchat_r path |
--condition-*, --receiver, --senders, --nichenet-* | NicheNet controls | used only by the nichenet_r path |
builtinCurrent OmicsClaw built-in communication path:
lianaCurrent OmicsClaw LIANA path:
cellphonedbCurrent OmicsClaw CellPhoneDB path:
species=human in the current wrappercellchat_rCurrent OmicsClaw CellChat path:
nichenet_rCurrent OmicsClaw NicheNet path:
nichenetr Seurat-wrapper style workflow in RImportant implementation note:
pvalue empty for compatibility with the shared table schema, so its n_significant count is not a formal statistical outputSuccessful runs write:
processed.h5adreport.mdresult.jsontables/lr_interactions.csvtables/top_interactions.csvtables/sender_receiver_summary.csvtables/group_role_summary.csvtables/pathway_summary.csvfigures/interaction_heatmap.pngfigures/top_interactions.pngfigures/source_target_bubble.pngfigures/group_role_summary.pngfigures/pathway_summary.pngtables/nichenet_ligand_activities.csv when --method nichenet_rtables/nichenet_ligand_target_links.csv when --method nichenet_rfigures/nichenet_top_ligands.png when --method nichenet_rtables/cellchat_pathways.csv, tables/cellchat_centrality.csv, tables/cellchat_count_matrix.csv, and tables/cellchat_weight_matrix.csv when --method cellchat_rtables/cellphonedb_means.csv, tables/cellphonedb_pvalues.csv, and tables/cellphonedb_significant_means.csv when --method cellphonedb returns themThe current wrapper writes direct figure outputs rather than a recipe-driven gallery:
figures/interaction_heatmap.pngfigures/top_interactions.pngfigures/source_target_bubble.pngfigures/group_role_summary.pngfigures/pathway_summary.pngfigures/nichenet_top_ligands.png for nichenet_rreport.mdfigures/interaction_heatmap.pngfigures/source_target_bubble.pngtables/top_interactions.csvtables/lr_interactions.csvprocessed.h5adpvalue empty because it does not run a formal significance test.cellphonedb is currently human-only in this wrapper.nichenet_r is currently human-only in this wrapper and requires explicit receiver / sender / condition metadata.cellchat_r requires a working R environment with CellChat, SingleCellExperiment, and zellkonverter through the shared bridge.README.md and notebook-style reproducibility artifacts when notebook export dependencies are available.cell_type_key as the most important scientific input because all backends aggregate by that grouping.pvalue values as formal significance statistics; for this backend they are intentionally left empty.cellphonedb is human-only in the current wrapper.knowledge_base/knowhows/KH-sc-cell-communication-guardrails.md.knowledge_base/skill-guides/singlecell/sc-cell-communication.md.| Flag | Type | Default | Description | Validation |
|---|---|---|---|---|
--input | str | — | Input .h5ad file | required unless --demo |
--output | str | — | Output directory | required |
--demo | flag | off | Run with bundled annotated demo data | — |
--method | str | builtin | Communication backend: builtin, liana, cellphonedb, cellchat_r, nichenet_r | validated against METHOD_REGISTRY |
--cell-type-key | str | cell_type | Cell type column in obs | — |
--species | str | human | Organism for L-R resource selection | choices: human, mouse |
--cellphonedb-counts-data | str | hgnc_symbol | Gene ID format for CellPhoneDB | choices: ensembl, gene_name, hgnc_symbol |
--cellphonedb-iterations | int | 1000 | Permutation iterations for CellPhoneDB | cellphonedb only |
--cellphonedb-threshold | float | 0.1 | Minimum expression fraction for CellPhoneDB | cellphonedb only |
--cellphonedb-threads | int | 4 | Parallel threads for CellPhoneDB | cellphonedb only |
--cellphonedb-pvalue | float | 0.05 | P-value cutoff for CellPhoneDB results | cellphonedb only |
--cellchat-prob-type | str | triMean | Aggregation method for CellChat | choices: triMean, truncatedMean, thresholdedMean, median |
--cellchat-min-cells | int | 10 | Minimum cells per cell type for CellChat | cellchat_r only |
--condition-key | str | condition | Condition column for NicheNet | nichenet_r only |
--condition-oi | str | stim | Condition of interest label | nichenet_r only |
--condition-ref | str | ctrl | Reference condition label | nichenet_r only |
--receiver | str | `` | Receiver cell type for NicheNet | nichenet_r only |
--senders | str | `` | Comma-separated sender cell types for NicheNet | nichenet_r only |
--nichenet-top-ligands | int | 20 | Number of top ligands to report | nichenet_r only |
--nichenet-expression-pct | float | 0.10 | Minimum expression fraction for ligand filtering | nichenet_r only |
--nichenet-lfc-cutoff | float | 0.25 | log2FC cutoff for receiver DE gene set | nichenet_r only |
--r-enhanced | flag | off | Also render R Enhanced ggplot2 figures | — |
Activated by --r-enhanced. Files written to figures/r_enhanced/.
| Renderer | Output file | figure_data CSV | Plot description | Required R packages |
|---|---|---|---|---|
plot_ccc_heatmap | r_ccc_heatmap.png | sender_receiver_summary.csv | Sender-receiver interaction strength heatmap | ggplot2 |
plot_ccc_network | r_ccc_network.png | sender_receiver_summary.csv | Network graph of cell-cell interactions | ggplot2, igraph |
plot_ccc_bubble | r_ccc_bubble.png | top_interactions.csv | Bubble plot of top ligand-receptor pairs | ggplot2 |
plot_ccc_stat_bar | r_ccc_stat_bar.png | group_role_summary.csv | Bar chart of interaction counts per cell type role | ggplot2 |
plot_ccc_stat_violin | r_ccc_stat_violin.png | top_interactions.csv | Violin plot of interaction scores per cell type | ggplot2 |
plot_ccc_stat_scatter | r_ccc_stat_scatter.png | top_interactions.csv | Scatter plot of L-R pair scores | ggplot2 |
plot_ccc_bipartite | r_ccc_bipartite.png | top_interactions.csv | Bipartite graph of sender-ligand-receptor-receiver | ggplot2, igraph |
plot_ccc_diff_network | r_ccc_diff_network.png | sender_receiver_summary.csv | Differential interaction network between conditions | ggplot2, igraph |
liana: falls back to builtin if the liana Python package is not installed. The fallback uses the built-in heuristic scorer (5 curated L-R pairs only). The report records both the requested and executed methods.cellchat_r: requires R environment with CellChat, SingleCellExperiment, and zellkonverter. Fails with an actionable error if the R stack is missing.nichenet_r: requires R environment with nichenetr, Seurat, and supporting packages. Human-only in the current wrapper.cellphonedb: human-only in the current wrapper. Requires cellphonedb Python package.Upstream: sc-clustering or sc-cell-annotation Downstream: Terminal analysis. Consider: sc-grn (gene regulatory networks)
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
