96 lines
3.1 KiB
Markdown
96 lines
3.1 KiB
Markdown
# AGENTS.md - bbav5 基因家族分析执行系统
|
|||
|
|
|
||
|
|
## 项目定位
|
||
|
|
|
||
|
|
**执行型 Agent** - 运行基因家族分析流程,调用 run_*.py 脚本
|
||
|
|
|
||
|
|
## 分析阶段速查
|
||
|
|
|
||
|
|
### 核心阶段 (13个)
|
||
|
|
|
||
|
|
| 阶段 | 脚本 | 功能 | 环境 |
|
||
|
|
|------|------|------|------|
|
||
|
|
| identify | run_identify.py | 基因鉴定 | test4py311 |
|
||
|
|
| phylogeny | run_phylogeny.py | 系统发育树 | test4py311 |
|
||
|
|
| motif | run_motif.py | MEME motif | test4py311 |
|
||
|
|
| domain | run_domain.py | Pfam结构域 | test4py311 |
|
||
|
|
| structure | run_structure.py | 基因结构 | test4py311 |
|
||
|
|
| promoter | run_promoter.py | 启动子分析 | test4py311 |
|
||
|
|
| synteny | run_synteny.py | 种内共线性 | test4py311 |
|
||
|
|
| interspecies_synteny | run_interspecies_synteny.py | 种间共线性 | test4py311 |
|
||
|
|
| karyotype | run_karyotype.py | 核型图 | test4py311 |
|
||
|
|
| combo | run_combo.py | 组合图 | test4py311 |
|
||
|
|
| ppi | run_ppi.py | 蛋白互作 | test4py311 |
|
||
|
|
| expression | run_expression.py | 表达分析 | **rnaR** |
|
||
|
|
| pep | run_pep.py | PEP分析 | test4py311 |
|
||
|
|
|
||
|
|
### RNA-seq 流水线 (5个)
|
||
|
|
|
||
|
|
| 阶段 | 脚本 | 功能 | 环境 |
|
||
|
|
|------|------|------|------|
|
||
|
|
| rnaseq_qc | run_rnaseq_qc.py | 质控 | **rnaseq** |
|
||
|
|
| rnaseq_align | run_rnaseq_align.py | 比对 | **rnaseq** |
|
||
|
|
| rnaseq_quant | run_rnaseq_quant.py | 定量 | **rnaseq** |
|
||
|
|
| rnaseq_de | run_rnaseq_de.py | 差异表达 | **rnaseq** |
|
||
|
|
| rnaseq_enrich | run_rnaseq_enrich_v2.py | 富集分析 | **rnaseq** |
|
||
|
|
|
||
|
|
### 辅助阶段 (9个,无契约)
|
||
|
|
|
||
|
|
genome_prepare_v2, kaks_v2, ortholog, cis_element_http, fetch_expression, tree_plot, ppi_v2, ppi_v2_with_repro, summary
|
||
|
|
|
||
|
|
## 执行流程
|
||
|
|
|
||
|
|
1. **识别意图** → 确定目标阶段
|
||
|
|
2. **检查输入** → 验证契约文件存在 (input/ 或 final/)
|
||
|
|
3. **执行脚本** → `python apps/cli/run_*.py` 或调用 Skills
|
||
|
|
4. **验证输出** → 检查 final/ 目录产物 + Manifest
|
||
|
|
5. **报告结果** → artifacts 列表 + metrics 摘要
|
||
|
|
|
||
|
|
## Conda 环境 [关键]
|
||
|
|
|
||
|
|
| 环境名 | 用途 | 激活命令 |
|
||
|
|
|--------|------|----------|
|
||
|
|
| test4py311 | 标准分析 | `conda activate test4py311` |
|
||
|
|
| rnaseq | RNA-seq流程 | `conda activate rnaseq` |
|
||
|
|
| rnaR | 表达绘图 | `conda activate rnaR` |
|
||
|
|
|
||
|
|
**注意**: expression阶段用rnaR,其他绘图用test4py311
|
||
|
|
|
||
|
|
## 关键目录
|
||
|
|
|
||
|
|
| 目录 | 内容 |
|
||
|
|
|------|------|
|
||
|
|
| `apps/cli/` | 27个 run_*.py 脚本 |
|
||
|
|
| `gfa_core/schemas/stages/` | 18个契约文件 *_contract_v1.md |
|
||
|
|
| `skills/` | 技能封装 |
|
||
|
|
| `scripts/` | R绘图脚本 |
|
||
|
|
|
||
|
|
## Manifest 结构
|
||
|
|
|
||
|
|
```json
|
||
|
|
{
|
||
|
|
"manifest_version": "1.0",
|
||
|
|
"stage": "identify",
|
||
|
|
"status": "success",
|
||
|
|
"prefix": "Ft.SPL",
|
||
|
|
"artifacts": { "key": {"path": "final/...", "type": "fasta"} },
|
||
|
|
"metrics": { ... }
|
||
|
|
}
|
||
|
|
```
|
||
|
|
|
||
|
|
## 工作原则
|
||
|
|
|
||
|
|
- **契约优先**: 有契约的阶段执行前检查输入,执行后验证输出
|
||
|
|
- **Manifest记录**: 每次执行生成完整 Manifest
|
||
|
|
- **可复现性**: reproducibility_v2.py 确保分析可追溯
|
||
|
|
- **错误诊断**: 失败时读错误日志,给具体修复建议
|
||
|
|
|
||
|
|
## 常用命令
|
||
|
|
|
||
|
|
```bash
|
||
|
|
# 运行鉴定阶段
|
||
|
|
python apps/cli/run_identify.py --prefix Ft.SPL --workspace ~/test_workspace
|
||
|
|
|
||
|
|
# 启动前端服务
|
||
|
|
python apps/web/vue3/main.py
|
||
|
|
```
|