← 返回 Skills 市场
bioinfo-style
作者
Fang, Chao
· GitHub ↗
· v1.0.0
· MIT-0
35
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install bioinfo-style
功能描述
标准化生物信息学分析流程。当用户提出以下任务时激活:(1) 测序数据分析(RNA-seq、ChIP-seq、WGS等),(2) 差异表达分析,(3) 通路/富集分析,(4) 组学数据可视化,(5) 统计检验,(6) 任何需要调用成熟生信工具而非编写Python/R算法代码的分析任务。核心原则:优先调用工具,记录s...
使用说明 (SKILL.md)
Bioinformatics Analysis - 标准化生信分析
核心原则
工具 > 代码:优先使用成熟的生信工具,绝不自己写算法代码。 过程记录:每一步都写 shell 脚本并执行,而非在对话中描述。 可重复性:所有分析必须可以通过脚本重现。
工作流程
1. 环境准备
# 读取 ~/.bashrc 中的环境配置
source ~/.bashrc
# 激活或创建 mamba 环境
mamba activate \x3Cenv_name> # 已有环境
mamba create -n \x3Cenv_name> -c \x3Cchannels> \x3Cpackages> # 创建新环境
# 常用生信环境示例
mamba create -n bioinfo -c conda-forge -c bioconda \
fastqc trimmomatic star featurecounts deseq2 \
samtools bedtools homer meme
2. 原始数据质量控制
# FastQC 质控
fastqc -o \x3Coutput_dir> -f fastq \x3Craw_data>/*.fastq.gz
# 批量处理脚本模板
cat > 01_qc.sh \x3C\x3C 'EOF'
#!/bin/bash
set -e
RAW_DIR="raw_data"
QC_DIR="qc_results"
mkdir -p $QC_DIR
for fq in $RAW_DIR/*.fastq.gz; do
echo "Processing: $fq"
fastqc -o $QC_DIR "$fq"
done
echo "QC completed"
EOF
bash 01_qc.sh
3. 序列比对/定量
# RNA-seq: STAR 比对 + featureCounts 定量
cat > 02_alignment.sh \x3C\x3C 'EOF'
#!/bin/bash
set -e
REF="genome/GRCh38"
SAMPLE="sample1"
GTF="annotation/gencode.v38.annotation.gtf"
mkdir -p alignment/$SAMPLE
star --genomeDir $REF \
--readFilesIn reads/$SAMPLE.fastq.gz \
--readFilesCommand zcat \
--outSAMtype BAM SortedByCoordinate \
--outFileNamePrefix alignment/$SAMPLE/
featureCounts -T 4 -t exon -g gene_id \
-a $GTF \
-o counts/$SAMPLE.counts \
alignment/$SAMPLE/Aligned.sortedByCoord.out.bam
EOF
bash 02_alignment.sh
4. 差异分析(Rmarkdown)
cat > 03_differential_analysis.Rmd \x3C\x3C 'EOF'
---
title: "差异表达分析报告"
author: "Bioinformatics Pipeline"
date: "`r Sys.Date()`"
output:
html_document:
toc: true
theme: united
---
```{r setup, message=FALSE}
library(DESeq2)
library(ggplot2)
library(pheatmap)
library(clusterProfiler)
# 读取计数矩阵
count_matrix \x3C- read.table("counts/matrix.txt", header=TRUE, row.names=1)
coldata \x3C- read.table("metadata.txt", header=TRUE, row.names=1)
dds \x3C- DESeqDataSetFromMatrix(countData=count_matrix,
colData=coldata,
design=~condition)
dds \x3C- DESeq(dds)
results \x3C- results(dds)
# MA plot
plotMA(results)
# 热图
top_genes \x3C- rownames(head(results[order(results$padj),], 20))
pheatmap(assay(vst(dds))[top_genes,])
EOF Rscript -e "rmarkdown::render('03_differential_analysis.Rmd')"
### 5. 富集分析
```bash
cat > 04_enrichment.sh \x3C\x3C 'EOF'
#!/bin/bash
set -e
GENELIST="differential_genes.txt"
OUTPUT="enrichment_results"
# GO富集
enrichment.sh $GENELIST BP CC MF BP
clusterProfiler --dotplot --gseGO
# KEGG通路
enrichment.sh $GENELIST KEGG
# Reactome
enrichment.sh $GENELIST Reactome
EOF
bash 04_enrichment.sh
工具优先级
- QC: FastQC, MultiQC, Trimmomatic, Cutadapt
- 比对: STAR, BWA, Bowtie2, HISAT2
- 定量: featureCounts, htseq-count, Salmon, kallisto
- 差异分析: DESeq2, edgeR, limma(仅用已有R包,不写算法)
- 富集分析: clusterProfiler, GSEA, Enrichr
- 可视化: ggplot2, pheatmap, Gviz, IGV
- 基因组操作: BEDTools, SAMtools, BCFtools
关键规则
- 绝不写Python脚本做已有的生信工具能做的事
- 每一步都写 .sh 脚本并执行,记录到
logs/目录 - 原始数据不修改,所有处理脚本化
- 输出结构标准化:
project/ ├── raw_data/ # 原始数据(只读) ├── qc_results/ # 质控结果 ├── alignment/ # 比对结果 ├── counts/ # 表达矩阵 ├── results/ # 差异分析结果 ├── figures/ # 图片 ├── scripts/ # 所有分析脚本 └── logs/ # 执行日志 - Rmarkdown 报告包含:方法描述、统计结果、可视化图表
常用工具速查
分析脚本模板
安全使用建议
Install this only if you expect the agent to run local bioinformatics commands, create scripts/logs/reports, and possibly create or update Conda/Mamba environments. Review generated commands before running them on valuable data or shared systems, and run analyses in an isolated project directory or environment when possible.
能力评估
Purpose & Capability
The skill's capabilities match its stated purpose: sequencing QC, alignment, quantification, differential expression, enrichment analysis, visualization, and reproducible reporting.
Instruction Scope
The activation text is broad within bioinformatics and tells the agent to script and execute each step, but that behavior is explicit and central to the advertised workflow rather than hidden or unrelated.
Install Mechanism
The skill includes Mamba/Conda environment creation and package installation examples, which can change local tool environments and use network/storage, but these are disclosed as setup steps for standard bioinformatics tools.
Credentials
Local shell scripts, R Markdown reports, logs, and project output directories are proportionate for reproducible bioinformatics analysis; no credential access, cloud upload, or unrelated data collection was found.
Persistence & Privilege
Persistence is limited to generated scripts, logs, analysis outputs, reports, and optional local tool environments; no background service, privilege escalation, or hidden long-running persistence was evident.
如何使用
- 确保已安装 OpenClaw(本地或 Docker 部署)
- 在对话框中输入安装命令:
/install bioinfo-style - 安装完成后,直接呼叫该 Skill 的名称或使用
/bioinfo-style触发 - 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
Initial release of standardized bioinformatics analysis workflow.
- Provides a reproducible pipeline for sequencing data analysis (QC, alignment, quantification, DE analysis, enrichment, visualization).
- Emphasizes use of established bioinformatics tools over custom code.
- Every step is scripted (shell/bash), with logs and organized output directories.
- Results and visualization are generated through Rmarkdown reports.
- Ensures raw data is unmodified and all processing is fully documented and repeatable.
- Templates and tool references included for easy adaptation.
元数据
常见问题
bioinfo-style 是什么?
标准化生物信息学分析流程。当用户提出以下任务时激活:(1) 测序数据分析(RNA-seq、ChIP-seq、WGS等),(2) 差异表达分析,(3) 通路/富集分析,(4) 组学数据可视化,(5) 统计检验,(6) 任何需要调用成熟生信工具而非编写Python/R算法代码的分析任务。核心原则:优先调用工具,记录s... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 35 次。
如何安装 bioinfo-style?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install bioinfo-style」即可一键安装,无需额外配置。
bioinfo-style 是免费的吗?
是的,bioinfo-style 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
bioinfo-style 支持哪些平台?
bioinfo-style 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。
谁开发了 bioinfo-style?
由 Fang, Chao(@biociao)开发并维护,当前版本 v1.0.0。
推荐 Skills