shiganluo/skill/bioinformatics-wiki

Bioinformatics Wiki Documentation

Use when creating or editing omics/bioinformatics technique wiki pages in the user's Obsidian vault. These live under Wiki/ScienceEngineering/DataScience/Bioinformatics/Omics/.

Reference Style

The best template is ATAC-seq.md in the Omics directory — read it before writing any new doc. MS.md is another good example (larger, covers qualitative + quantitative + PTM + clinical).

Content Philosophy

User explicitly corrected twice: 拒绝教科书式写法. Rules:

• Do NOT list every technical parameter exhaustively
• Organize by practical scenario / clinical application, not by encyclopedia structure
• Emphasize feasibility (可行性) and real-world applicability (落地性)
• Include workflow diagrams, comparison tables, and tool recommendations — not dry definitions

Document Structure

Frontmatter

---
title: 技术名称（中文全称+英文缩写）
tags: [大类, 子类, 组学]
wiki: https://en.wikipedia.org/wiki/Article
---

Standard Sections (adapt as needed)

1. 一、概述 — Definition, inventor/year, position in omics landscape
2. 二、基本原理 — Core concepts, ASCII art diagrams, technology comparison table
3. 三、样品前处理/实验流程 — Step-by-step ASCII tree (│ ├── └──)
4. 四、数据采集模式 — Mode comparison table (if applicable)
5. 五、数据分析流程 — Standard pipeline flowchart + QC metrics table
6. 六、下游分析 — Differential analysis, enrichment, visualization
7. 七、工具与平台 — Tool table (tool | purpose | notes)
8. 八、应用领域 — Organized by scenario/disease
9. 九、参考资源 — Original papers + tutorial links

Formatting Conventions

• Chinese prose, English technical terms kept as-is
• Flow diagrams: ASCII art with │ ├── └── and → for outputs
• Comparisons: Markdown tables
• Tool lists: tables (Tool | Use | Notes)
• QC metrics: tables (Metric | Expected | Description)

Renumbering Pitfall

When inserting a new section mid-document, ALL subsequent section numbers (e.g. 七→八→九) AND sub-section numbers (e.g. 7.1→8.1) must cascade-update.

Safe approach: use execute_code with hermes_tools.patch() in a loop. Build the mapping first (old→new for both sections and sub-sections), then apply. Change sub-section numbers before section headers to avoid conflicts.

Sample Metadata Descriptions

When user pastes a GEO/SRA sample table and asks for a description ("添加描述"), see references/sample-metadata-descriptions.md for conventions. Key rules:

• Always explain relationships between experimental factors, not just list them (user corrected: "PKR和Dicer的关系没有交代")
• When user asks to verify design ("确保真实设计"), check SRA pages (accessible, rich metadata) before GEO (reCAPTCHA blocked). CNGB metadata (CNP/CNS accessions) provides project-level context if user supplies it.
• When multiple datasets are in conversation, "上面的样本" likely means the most recent one unless explicitly stated otherwise.

Result Directory Scanning

When analyzing a research group's results folder structure (not writing wiki pages, but constructing research narratives from existing data), see references/result-directory-conventions.md for:

• Directory name → analysis type mapping (GSEA, TE, DDR, SNP, etc.)
• How to extract experimental conditions from folder/file names
• How to cluster projects into a coherent thesis narrative
• Common pitfalls (typos in gene names, lock files, compressed archives)

Existing Documents