tooluniverse-spatial-omics-analysis

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Spatial Multi-Omics Analysis Pipeline

空间多组学分析流程

Comprehensive biological interpretation of spatial omics data. Transforms spatially variable genes (SVGs), domain annotations, and tissue context into actionable biological insights covering pathway enrichment, cell-cell interactions, druggable targets, immune microenvironment, and multi-modal integration.

KEY PRINCIPLES:

Report-first approach - Create report file FIRST, then populate progressively
Domain-by-domain analysis - Characterize each spatial region independently before comparison
Gene-list-centric - Analyze user-provided SVGs and marker genes with ToolUniverse databases
Biological interpretation - Go beyond statistics to explain biological meaning of spatial patterns
Disease focus - Emphasize disease mechanisms and therapeutic opportunities when disease context is provided
Evidence grading - Grade all evidence as T1 (human/clinical) to T4 (computational)
Multi-modal thinking - Integrate RNA, protein, and metabolite information when available
Validation guidance - Suggest experimental validation approaches for key findings
Source references - Every statement must cite tool/database source
Completeness checklist - Mandatory section showing analysis coverage
English-first queries - Always use English terms in tool calls. Respond in user's language

对空间多组学数据进行全面的生物学解读。将空间可变基因（SVG）、区域注释和组织背景转化为可落地的生物学洞见，涵盖通路富集、细胞间互作、可成药靶点、免疫微环境以及多模态整合等内容。

核心原则:

报告优先方法 - 先创建报告文件，再逐步填充内容
逐区域分析 - 在进行区域间比较前，先独立表征每个空间区域
基因列表为中心 - 结合ToolUniverse数据库分析用户提供的SVG和标记基因
生物学解读 - 超越统计层面，解释空间模式的生物学意义
聚焦疾病 - 当提供疾病背景时，重点关注疾病机制和治疗机会
证据分级 - 将所有证据分为T1（人类/临床）至T4（计算预测）四个等级
多模态思维 - 若有可用数据，整合RNA、蛋白质和代谢物信息
验证指导 - 为关键发现提供实验验证方法建议
来源引用 - 所有结论必须标注工具/数据库来源
完整性检查清单 - 必须包含显示分析覆盖范围的章节
英文优先查询 - 在工具调用中始终使用英文术语，以用户使用的语言回复

When to Use This Skill

何时使用该技能

Apply when users:

Provide spatially variable genes from spatial transcriptomics experiments
Ask about biological interpretation of spatial domains/clusters
Need pathway enrichment analysis of spatial gene expression data
Want to understand cell-cell interactions from spatial data
Ask about tumor microenvironment heterogeneity from spatial omics
Need druggable targets in specific spatial regions
Ask about tissue zonation patterns (liver, brain, kidney)
Want to integrate spatial transcriptomics + proteomics data
Ask about immune infiltration patterns from spatial data
Need to compare healthy vs disease regions spatially
Ask "What pathways are enriched in this tumor core vs tumor margin?"
Ask "What cell-cell interactions occur in this spatial domain?"

NOT for (use other skills instead):

Single gene interpretation without spatial context -> Use
```
tooluniverse-target-research
```
Variant interpretation -> Use
```
tooluniverse-variant-interpretation
```
Drug safety profiling -> Use
```
tooluniverse-adverse-event-detection
```
Disease-only analysis without spatial data -> Use
```
tooluniverse-multiomic-disease-characterization
```
GWAS analysis -> Use
```
tooluniverse-gwas-*
```
skills
Bulk RNA-seq (non-spatial) -> Use
```
tooluniverse-systems-biology
```

当用户出现以下情况时适用：

提供空间转录组实验得到的空间可变基因
询问空间区域/聚类的生物学解读
需要对空间基因表达数据进行通路富集分析
希望从空间数据中理解细胞间互作
询问空间多组学中的肿瘤微环境异质性
需要识别特定空间区域中的可成药靶点
询问组织分区模式（肝脏、大脑、肾脏）
希望整合空间转录组与蛋白质组数据
询问空间数据中的免疫浸润模式
需要对健康与疾病区域进行空间层面的比较
提问“肿瘤核心区与边缘区的富集通路有哪些？”
提问“该空间区域中存在哪些细胞间互作？”

不适用场景（请使用其他技能）：

无空间背景的单基因解读 -> 使用
```
tooluniverse-target-research
```
变异解读 -> 使用
```
tooluniverse-variant-interpretation
```
药物安全性分析 -> 使用
```
tooluniverse-adverse-event-detection
```

无空间数据的纯疾病分析 -> 使用

tooluniverse-multiomic-disease-characterization

GWAS分析 -> 使用
```
tooluniverse-gwas-*
```
系列技能
bulk RNA-seq（非空间）分析 -> 使用
```
tooluniverse-systems-biology
```

Input Parameters

输入参数

Parameter	Required	Description	Example
svgs	Yes	Spatially variable genes (gene symbols)	`['EGFR', 'CDH1', 'VIM', 'MYC', 'CD3E']`
tissue_type	Yes	Tissue/organ type	`brain` , `liver` , `lung` , `breast` , `skin`
technology	No	Spatial omics platform used	`10x Visium` , `MERFISH` , `DBiTplus` , `SLIDE-seq`
disease_context	No	Disease if applicable	`breast cancer` , `Alzheimer disease` , `liver cirrhosis`
spatial_domains	No	Dict mapping domain name to marker genes	`{'Tumor core': ['MYC','EGFR'], 'Stroma': ['VIM','COL1A1']}`
cell_types	No	Cell types identified in deconvolution	`['Epithelial', 'T cell', 'Macrophage', 'Fibroblast']`
proteins	No	Proteins detected (if multi-modal)	`['CD3', 'CD8', 'PD-L1', 'Ki67']`
metabolites	No	Metabolites detected (if SpatialMETA)	`['glutamine', 'lactate', 'ATP']`

参数	是否必填	描述	示例
svgs	是	空间可变基因（基因符号）	`['EGFR', 'CDH1', 'VIM', 'MYC', 'CD3E']`
tissue_type	是	组织/器官类型	`brain` , `liver` , `lung` , `breast` , `skin`
technology	否	所用的空间多组学平台	`10x Visium` , `MERFISH` , `DBiTplus` , `SLIDE-seq`
disease_context	否	适用的疾病（如有）	`breast cancer` , `Alzheimer disease` , `liver cirrhosis`
spatial_domains	否	映射区域名称与标记基因的字典	`{'Tumor core': ['MYC','EGFR'], 'Stroma': ['VIM','COL1A1']}`
cell_types	否	去卷积识别出的细胞类型	`['Epithelial', 'T cell', 'Macrophage', 'Fibroblast']`
proteins	否	检测到的蛋白质（多模态数据时）	`['CD3', 'CD8', 'PD-L1', 'Ki67']`
metabolites	否	检测到的代谢物（SpatialMETA数据时）	`['glutamine', 'lactate', 'ATP']`

Spatial Omics Integration Score (0-100)

空间多组学整合评分（0-100）

Score Components

评分组成

Data Completeness (0-30 points):

SVGs provided (>10 genes): 5 points
Disease context provided: 5 points
Spatial domains defined: 5 points
Cell type composition available: 5 points
Multi-modal data (protein/metabolite): 5 points
Literature context found: 5 points

Biological Insight (0-40 points):

Significant pathway enrichment (FDR < 0.05): 10 points
Cell-cell interaction predictions: 10 points
Disease mechanism identified: 10 points
Druggable targets found in disease regions: 10 points

Evidence Quality (0-30 points):

Cross-database validation (gene found in 3+ databases): 10 points
Clinical validation (approved drugs for spatial targets): 10 points
Literature support (PubMed evidence for spatial patterns): 10 points

数据完整性（0-30分）:

提供SVG（>10个基因）：5分
提供疾病背景：5分
定义空间区域：5分
提供细胞类型组成：5分
提供多模态数据（蛋白质/代谢物）：5分
找到文献背景：5分

生物学洞见（0-40分）:

显著通路富集（FDR < 0.05）：10分
细胞间互作预测：10分
识别疾病机制：10分
在疾病区域找到可成药靶点：10分

证据质量（0-30分）:

跨数据库验证（基因在3个及以上数据库中存在）：10分
临床验证（针对空间靶点的已获批药物）：10分
文献支持（PubMed中存在空间模式的证据）：10分

Score Interpretation

评分解读

Score	Tier	Interpretation
80-100	Excellent	Comprehensive spatial characterization, strong biological insights, druggable targets identified
60-79	Good	Good pathway and interaction analysis, some disease/therapeutic context
40-59	Moderate	Basic enrichment complete, limited spatial domain comparison or interaction analysis
0-39	Limited	Minimal data, gene-level annotation only

分数	等级	解读
80-100	优秀	全面的空间表征，丰富的生物学洞见，已识别可成药靶点
60-79	良好	通路与互作分析质量佳，具备部分疾病/治疗背景
40-59	中等	完成基础富集分析，空间区域比较或互作分析有限
0-39	有限	数据量极少，仅完成基因层面注释

Evidence Grading System

证据分级体系

Tier	Symbol	Criteria	Examples
T1	[T1]	Direct human evidence, clinical proof	FDA-approved drug for spatial target, validated biomarker
T2	[T2]	Experimental evidence	Validated spatial pattern in literature, known ligand-receptor pair
T3	[T3]	Computational/database evidence	PPI network prediction, pathway enrichment, expression correlation
T4	[T4]	Annotation/prediction only	GO annotation, text-mined association, predicted interaction

等级	符号	标准	示例
T1	[T1]	直接人类证据，临床验证	针对空间靶点的FDA获批药物，经过验证的生物标志物
T2	[T2]	实验证据	文献中已验证的空间模式，已知的配体-受体对
T3	[T3]	计算/数据库证据	PPI网络预测，通路富集，表达相关性
T4	[T4]	仅注释/预测	GO注释，文本挖掘关联，预测的互作

Report Template

报告模板

Create this file structure at the start:

{tissue}_{disease}_spatial_omics_report.md

markdown

undefined

在开始时创建以下文件结构：

{tissue}_{disease}_spatial_omics_report.md

markdown

undefined

Spatial Multi-Omics Analysis Report: {Tissue Type}

空间多组学分析报告：{组织类型}

Report Generated: {date} Technology: {platform} Tissue: {tissue_type} Disease Context: {disease or "Normal tissue"} Total SVGs Analyzed: {count} Spatial Domains: {count} Spatial Omics Integration Score: (to be calculated)

报告生成时间: {日期} 技术平台: {平台} 组织: {tissue_type} 疾病背景: {疾病或“正常组织”} 分析的SVG总数: {数量} 空间区域数量: {数量} 空间多组学整合评分: （待计算）

Executive Summary

执行摘要

(2-3 sentence synthesis of key spatial findings - fill after all phases complete)

（2-3句话总结关键空间发现 - 完成所有阶段后填写）

1. Tissue & Disease Context

1. 组织与疾病背景

Tissue Information

组织信息

Property	Value	Source
Tissue type
Disease
Expected cell types		HPA

属性	数值	来源
组织类型
疾病
预期细胞类型		HPA

Disease Identifiers (if applicable)

疾病标识符（如适用）

System	ID	Source

Sources: (tools used)

系统	ID	来源

来源: （使用的工具）

2. Spatially Variable Gene Characterization

2. 空间可变基因表征

2.1 Gene ID Resolution

2.1 基因ID解析

Gene Symbol	Ensembl ID	Entrez ID	UniProt	Function	Source

基因符号	Ensembl ID	Entrez ID	UniProt	功能	来源

2.2 Tissue Expression Patterns

2.2 组织表达模式

Gene	Tissue Expression	Specificity	Source

基因	组织表达情况	特异性	来源

2.3 Subcellular Localization

2.3 亚细胞定位

Gene	Location	Confidence	Source

基因	定位	置信度	来源

2.4 Disease Associations

2.4 疾病关联

Gene	Disease	Score	Evidence	Source

Sources: (tools used)

基因	疾病	评分	证据	来源

来源: （使用的工具）

3. Pathway Enrichment Analysis

3. 通路富集分析

3.1 STRING Functional Enrichment

3.1 STRING功能富集

Category	Term	Description	P-value	FDR	Genes	Source

类别	术语	描述	P值	FDR	基因	来源

3.2 Reactome Pathway Analysis

3.2 Reactome通路分析

Pathway ID	Name	P-value	FDR	Genes Found	Total Genes	Source

通路ID	名称	P值	FDR	找到的基因	总基因数	来源

3.3 GO Biological Processes

3.3 GO生物过程

GO Term	Description	P-value	FDR	Genes	Source

GO术语	描述	P值	FDR	基因	来源

3.4 GO Molecular Functions

3.4 GO分子功能

GO Term	Description	P-value	FDR	Genes	Source

GO术语	描述	P值	FDR	基因	来源

3.5 GO Cellular Components

3.5 GO细胞组分

GO Term	Description	P-value	FDR	Genes	Source

GO术语	描述	P值	FDR	基因	来源

Pathway Summary

通路总结

Top enriched pathways:
Key biological processes:
Spatial pathway implications:

Sources: (tools used)

排名靠前的富集通路:
关键生物过程:
空间通路意义:

来源: （使用的工具）

4. Spatial Domain Characterization

4. 空间区域表征

Domain: {domain_name}

区域: {区域名称}

Marker Genes

标记基因

Gene	Function	Pathways	Source

基因	功能	通路	来源

Enriched Pathways (domain-specific)

富集通路（区域特异性）

Pathway	P-value	FDR	Genes	Source

通路	P值	FDR	基因	来源

Cell Type Signature

细胞类型特征

Cell Type	Marker Genes Present	Confidence

细胞类型	存在的标记基因	置信度

Biological Interpretation

生物学解读

(Narrative interpretation of this domain)

(Repeat for each domain)

（该区域的描述性解读）

（每个区域重复上述内容）

4.N Domain Comparison

4.N 区域比较

Feature	Domain 1	Domain 2	Domain 3
Top pathway
Cell types
Disease relevance

Sources: (tools used)

特征	区域1	区域2	区域3
顶级通路
细胞类型
疾病相关性

来源: （使用的工具）

5. Cell-Cell Interaction Inference

5. 细胞间互作推断

5.1 Protein-Protein Interactions (STRING)

5.1 蛋白质-蛋白质互作（STRING）

Protein A	Protein B	Score	Type	Source

蛋白质A	蛋白质B	评分	类型	来源

5.2 Ligand-Receptor Pairs

5.2 配体-受体对

Ligand	Receptor	Domain (Ligand)	Domain (Receptor)	Evidence	Source

配体	受体	配体所在区域	受体所在区域	证据	来源

5.3 Signaling Pathways

5.3 信号通路

Pathway	Components in Data	Spatial Distribution	Source

通路	数据中存在的组分	空间分布	来源

5.4 Interaction Network Summary

5.4 互作网络总结

Key interaction hubs:
Cross-domain interactions:
Predicted cell-cell communication axes:

Sources: (tools used)

关键互作枢纽:
跨区域互作:
预测的细胞间通讯轴:

来源: （使用的工具）

6. Disease & Therapeutic Context

6. 疾病与治疗背景

6.1 Disease Gene Overlap

6.1 疾病基因重叠

Gene	Disease Association Score	Evidence Type	Source

基因	疾病关联评分	证据类型	来源

6.2 Druggable Targets in Spatial Domains

6.2 空间区域中的可成药靶点

Gene	Domain	Tractability	Modality	Approved Drugs	Source

基因	区域	成药性	模态	已获批药物	来源

6.3 Drug Mechanisms Relevant to Spatial Targets

6.3 与空间靶点相关的药物机制

Drug	Target	Mechanism	Phase	Source

药物	靶点	机制	研发阶段	来源

6.4 Clinical Trials

6.4 临床试验

NCT ID	Title	Target Gene	Phase	Status	Source

NCT ID	标题	靶向基因	阶段	状态	来源

Therapeutic Summary

治疗总结

Druggable genes in disease regions:
Approved therapies:
Pipeline drugs:
Novel opportunities:

Sources: (tools used)

疾病区域中的可成药基因:
已获批疗法:
在研药物:
新机遇:

来源: （使用的工具）

7. Multi-Modal Integration

7. 多模态整合

7.1 Protein-RNA Concordance (if protein data available)

7.1 蛋白质-RNA一致性（若有蛋白质数据）

Gene/Protein	RNA Pattern	Protein Pattern	Concordance	Source

基因/蛋白质	RNA模式	蛋白质模式	一致性	来源

7.2 Subcellular Context

7.2 亚细胞背景

Gene	mRNA Location (spatial)	Protein Location (HPA)	Concordance	Source

基因	RNA空间定位	蛋白质定位（HPA）	一致性	来源

7.3 Metabolic Context (if metabolomics available)

7.3 代谢背景（若有代谢组数据）

Gene	Metabolic Pathway	Metabolites Detected	Spatial Pattern	Source

Sources: (tools used)

基因	代谢通路	检测到的代谢物	空间模式	来源

来源: （使用的工具）

8. Immune Microenvironment (if relevant)

8. 免疫微环境（如适用）

8.1 Immune Cell Markers

8.1 免疫细胞标记

Cell Type	Marker Genes	Spatial Domain	Source

细胞类型	标记基因	空间区域	来源

8.2 Immune Checkpoint Expression

8.2 免疫检查点表达

Checkpoint	Gene	Expression Pattern	Source

检查点	基因	表达模式	来源

8.3 Tumor-Immune Interface (if cancer)

8.3 肿瘤-免疫界面（若为癌症）

Feature	Finding	Evidence	Source

特征	发现	证据	来源

Immune Summary

免疫总结

Immune infiltration pattern:
Key immune checkpoints:
Immunotherapy implications:

Sources: (tools used)

免疫浸润模式:
关键免疫检查点:
免疫治疗意义:

来源: （使用的工具）

9. Literature & Validation Context

9. 文献与验证背景

9.1 Literature Evidence

9.1 文献证据

PMID	Title	Relevance	Year	Source

PMID	标题	相关性	年份	来源

9.2 Known Spatial Patterns

9.2 已知空间模式

(Known tissue architecture/zonation from literature)

（文献中已知的组织架构/分区）

9.3 Validation Recommendations

9.3 验证建议

Priority	Gene/Target	Method	Rationale
High		IHC / smFISH
Medium		IF / ISH

Sources: (tools used)

优先级	基因/靶点	方法	理由
高		IHC / smFISH
中		IF / ISH

来源: （使用的工具）

Spatial Omics Integration Score

空间多组学整合评分

Component	Points	Max	Details
SVGs provided		5
Disease context		5
Spatial domains		5
Cell types		5
Multi-modal data		5
Literature context		5
Pathway enrichment		10
Cell-cell interactions		10
Disease mechanism		10
Druggable targets		10
Cross-database validation		10
Clinical validation		10
Literature support		10
TOTAL		100

Score: XX/100 - [Tier]

组成部分	得分	满分	详情
提供SVG		5
提供疾病背景		5
定义空间区域		5
提供细胞类型		5
提供多模态数据		5
找到文献背景		5
通路富集		10
细胞间互作		10
疾病机制		10
可成药靶点		10
跨数据库验证		10
临床验证		10
文献支持		10
总分		100

评分: XX/100 - [等级]

Completeness Checklist

完整性检查清单

References

参考文献

Data Sources Used

使用的数据源

#	Tool	Parameters	Section	Items Retrieved

#	工具	参数	章节	检索到的条目

Database Versions

数据库版本

OpenTargets: (current)
STRING: v12.0
Reactome: (current)
HPA: (current)
GTEx: v10

---

OpenTargets: （当前版本）
STRING: v12.0
Reactome: （当前版本）
HPA: （当前版本）
GTEx: v10

---

Phase 0: Input Processing & Disambiguation (ALWAYS FIRST)

阶段0：输入处理与消歧（始终优先执行）

Objective: Parse user input, resolve tissue/disease identifiers, establish analysis context.

目标: 解析用户输入，解析组织/疾病标识符，建立分析背景。

Tools Used

使用的工具

OpenTargets_get_disease_id_description_by_name (if disease context provided):

Input:
```
diseaseName
```
(string) - Disease name

Output:

{data: {search: {hits: [{id, name, description}]}}}

Use: Get MONDO/EFO IDs for disease queries

OpenTargets_get_disease_description_by_efoId:

Input:
```
efoId
```
(string) - Disease ID (e.g.,
```
MONDO_0007254
```
)

Output:

{data: {disease: {id, name, description, dbXRefs}}}

Use: Get full disease description

HPA_search_genes_by_query (tissue cell type context):

Input:
```
query
```
(string) - Search term
Output: List of gene entries matching query
Use: Verify tissue-relevant genes

OpenTargets_get_disease_id_description_by_name（若提供疾病背景）:

输入:
```
diseaseName
```
（字符串）- 疾病名称

输出:

{data: {search: {hits: [{id, name, description}]}}}

用途: 为疾病查询获取MONDO/EFO ID

OpenTargets_get_disease_description_by_efoId:

输入:
```
efoId
```
（字符串）- 疾病ID（例如
```
MONDO_0007254
```
）

输出:

{data: {disease: {id, name, description, dbXRefs}}}

用途: 获取完整的疾病描述

HPA_search_genes_by_query（组织细胞类型背景）:

输入:
```
query
```
（字符串）- 搜索词
输出: 匹配查询的基因条目列表
用途: 验证与组织相关的基因

Workflow

工作流程

Parse SVG list from user input (ensure valid gene symbols)
Identify tissue type and map to standard ontology term
If disease provided, resolve to MONDO/EFO ID using OpenTargets
Get disease description and cross-references
Determine analysis scope:
- Cancer? -> Include immune microenvironment, somatic mutations, druggable targets
- Neurological? -> Include brain region specificity, neuronal markers
- Metabolic? -> Include metabolic zonation, enzyme distribution
- Normal tissue? -> Focus on tissue architecture and cell type composition
Set up report file with header information

从用户输入中解析SVG列表（确保为有效的基因符号）
识别组织类型并映射到标准本体术语
若提供疾病信息，使用OpenTargets解析为MONDO/EFO ID
获取疾病描述和交叉引用
确定分析范围:
- 是否为癌症？-> 纳入免疫微环境、体细胞突变、可成药靶点分析
- 是否为神经系统疾病？-> 纳入脑区特异性、神经元标记分析
- 是否为代谢性疾病？-> 纳入代谢分区、酶分布分析
- 是否为正常组织？-> 聚焦组织架构和细胞类型组成
建立包含头部信息的报告文件

Decision Logic

决策逻辑

Cancer tissue: Enable immune microenvironment phase, CIViC/cBioPortal queries, immuno-oncology analysis
Normal tissue: Skip disease phases, focus on tissue zonation and cell type composition
Liver/kidney/brain: Enable zonation-specific analysis
No disease context: Proceed with tissue biology only
Small gene list (<20): Warn about limited enrichment power, emphasize gene-level analysis
Large gene list (>500): Suggest filtering to top SVGs by significance before enrichment

癌症组织: 启用免疫微环境阶段、CIViC/cBioPortal查询、肿瘤免疫分析
正常组织: 跳过疾病相关阶段，聚焦组织分区和细胞类型组成
肝脏/肾脏/大脑: 启用分区特异性分析
无疾病背景: 仅针对组织生物学进行分析
基因列表较小（<20个）: 提示富集能力有限，强调基因层面分析
基因列表较大（>500个）: 建议按显著性筛选出排名靠前的SVG后再进行富集

Phase 1: Gene Characterization

阶段1：基因表征

Objective: Resolve gene identifiers, annotate functions, tissue specificity, and subcellular localization.

目标: 解析基因标识符，注释功能、组织特异性和亚细胞定位。

Tools Used

使用的工具

MyGene_query_genes (gene ID resolution):

Input:
```
query
```
(string) - Gene symbol

Output:

{hits: [{_id, symbol, name, ensembl: {gene}, entrezgene}]}

Use: Resolve gene symbol to Ensembl ID, Entrez ID
NOTE: First hit may not be exact match - filter by
```
symbol
```
field

UniProt_get_function_by_accession (gene function):

Input:
```
accession
```
(string) - UniProt accession
Output: List of function description strings
Use: Get protein function annotation

UniProt_get_subcellular_location_by_accession (protein localization):

Input:
```
accession
```
(string)
Output: Subcellular location information
Use: Where the protein is located in the cell

HPA_get_subcellular_location (validated localization):

Input:
```
gene_name
```
(string) - Gene symbol

Output:

{gene_name, main_locations: [], additional_locations: [], location_summary}

Use: Experimentally validated protein subcellular location

HPA_get_rna_expression_by_source (tissue expression):

Input:
```
gene_name
```
(string),
```
source_type
```
(string: 'tissue'),
```
source_name
```
(string)

Output:

{data: {gene_name, source_type, source_name, expression_value, expression_level}}

Use: Check expression in the specific tissue of interest
NOTE: All 3 parameters are REQUIRED

HPA_get_comprehensive_gene_details_by_ensembl_id (full HPA data):

Input:

ensembl_id

(string),

include_isoforms

(bool),

include_images

(bool),

include_antibodies

(bool),

include_expression

(bool) - ALL 5 parameters REQUIRED

Output:

{ensembl_id, gene_name, uniprot_ids, summary, protein_classes, tissue_expression, cell_line_expression, ...}

Use: One-stop gene characterization from HPA
NOTE: Use
```
include_expression=True
```
for tissue data; set others to
```
False
```
for faster response

HPA_get_cancer_prognostics_by_gene (cancer prognosis):

Input:
```
ensembl_id
```
(string) - Ensembl gene ID (NOT gene_name)

Output:

{gene_name, prognostic_cancers_count, prognostic_summary: [{cancer_type, prognostic_type, p_value}]}

Use: Prognostic significance in cancer (if cancer context)

UniProtIDMap_gene_to_uniprot (ID mapping):

Input:
```
gene_name
```
(string),
```
organism
```
(string, default 'human')
Output: UniProt accession for the gene
Use: Map gene symbol to UniProt accession

MyGene_query_genes（基因ID解析）:

输入:
```
query
```
（字符串）- 基因符号

输出:

{hits: [{_id, symbol, name, ensembl: {gene}, entrezgene}]}

用途: 将基因符号解析为Ensembl ID、Entrez ID
注意: 第一个结果可能不是精确匹配 - 需按
```
symbol
```
字段过滤

UniProt_get_function_by_accession（基因功能）:

输入:
```
accession
```
（字符串）- UniProt登录号
输出: 功能描述字符串列表
用途: 获取蛋白质功能注释

UniProt_get_subcellular_location_by_accession（蛋白质定位）:

输入:
```
accession
```
（字符串）
输出: 亚细胞定位信息
用途: 确定蛋白质在细胞中的位置

HPA_get_subcellular_location（已验证的定位）:

输入:
```
gene_name
```
（字符串）- 基因符号

输出:

{gene_name, main_locations: [], additional_locations: [], location_summary}

用途: 获取经实验验证的蛋白质亚细胞定位

HPA_get_rna_expression_by_source（组织表达）:

输入:
```
gene_name
```
,
```
source_type
```
（字符串）,
```
source_name
```
（字符串）

输出:

{data: {gene_name, source_type, source_name, expression_value, expression_level}}

用途: 检查目标组织中的表达情况
注意: 所有3个参数均为必填项

HPA_get_comprehensive_gene_details_by_ensembl_id（完整HPA数据）:

输入:
```
ensembl_id
```
（字符串）,
```
include_isoforms
```
（布尔值）,
```
include_images
```
（布尔值）,
```
include_antibodies
```
（布尔值）,
```
include_expression
```
（布尔值）- 所有5个参数均为必填项

输出:

{ensembl_id, gene_name, uniprot_ids, summary, protein_classes, tissue_expression, cell_line_expression, ...}

用途: 从HPA一站式获取基因表征数据
注意: 若需要组织数据，设置
```
include_expression=True
```
；其他参数设为
```
False
```
以加快响应速度

HPA_get_cancer_prognostics_by_gene（癌症预后）:

输入:
```
ensembl_id
```
（字符串）- Ensembl基因ID（非基因名称）

输出:

{gene_name, prognostic_cancers_count, prognostic_summary: [{cancer_type, prognostic_type, p_value}]}

用途: 分析基因在癌症中的预后意义（若有癌症背景）

UniProtIDMap_gene_to_uniprot（ID映射）:

输入:
```
gene_name
```
（字符串）,
```
organism
```
（字符串，默认值'human'）
输出: 基因对应的UniProt登录号
用途: 将基因符号映射为UniProt登录号

Workflow

工作流程

For each SVG (batch if >20, sample top genes): a. Query MyGene to get Ensembl ID, Entrez ID b. Map to UniProt accession c. Get subcellular location from HPA d. Get tissue expression from HPA e. If cancer: check cancer prognostics
Compile gene characterization table
Identify genes with tissue-specific expression
Note genes with nuclear vs membrane vs secreted localization (relevant for spatial patterns)

针对每个SVG（若>20个则批量处理，选取排名靠前的基因）: a. 查询MyGene获取Ensembl ID、Entrez ID b. 映射为UniProt登录号 c. 从HPA获取亚细胞定位 d. 从HPA获取组织表达情况 e. 若为癌症背景：检查癌症预后意义
整理基因表征表格
识别具有组织特异性表达的基因
记录基因的核定位、膜定位或分泌定位（与空间模式相关）

Batch Strategy for Large Gene Lists

大基因列表的批量策略

10-50 genes: Characterize all individually
50-200 genes: Characterize top 50 by priority (known disease genes first), summarize rest
200+ genes: Characterize top 30, use enrichment for the full list
Always run pathway enrichment on the FULL list regardless

10-50个基因: 逐个表征所有基因
50-200个基因: 优先表征排名前50的基因（已知疾病基因优先），总结其余基因
200+个基因: 表征排名前30的基因，对完整列表进行富集分析
无论基因数量多少，始终对完整列表进行通路富集分析

Phase 2: Pathway & Functional Enrichment

阶段2：通路与功能富集

Objective: Identify biological pathways and functions enriched in SVGs and per-domain gene sets.

目标: 识别SVG和各区域基因集中富集的生物学通路与功能。

Tools Used

使用的工具

STRING_functional_enrichment (primary enrichment):

Input:
```
protein_ids
```
(array of gene symbols),
```
species
```
(int, 9606 for human)

Output:

{status: 'success', data: [{category, term, number_of_genes, number_of_genes_in_background, p_value, fdr, description, inputGenes, preferredNames}]}

Use: Comprehensive enrichment across GO, KEGG, Reactome, COMPARTMENTS, DISEASES

Categories:

Process

(GO:BP),

Function

(GO:MF),

Component

(GO:CC),

KEGG

Reactome

COMPARTMENTS

DISEASES

Keyword

PMID

NOTE: This is the PRIMARY enrichment tool. Returns all categories in one call

ReactomeAnalysis_pathway_enrichment (Reactome-specific):

Input:
```
identifiers
```
(string, space-separated gene symbols, NOT array)

Output:

{data: {token, pathways_found, pathways: [{pathway_id, name, p_value, fdr, entities_found, entities_total}]}}

Use: Detailed Reactome pathway analysis with hierarchy
NOTE: identifiers is a SPACE-SEPARATED STRING, not array

Reactome_map_uniprot_to_pathways (individual gene):

Input:
```
id
```
(string) - UniProt accession
Output: Plain list of pathway objects (no data wrapper)
Use: Map individual proteins to Reactome pathways

GO_get_annotations_for_gene (individual gene GO):

Input:
```
gene_id
```
(string) - Gene symbol or ID
Output: Plain list of GO annotation objects
Use: Get GO annotations for individual genes

kegg_search_pathway (KEGG pathway search):

Input:
```
query
```
(string) - Pathway name or keyword
Output: Pathway search results
Use: Find KEGG pathways relevant to spatial findings

WikiPathways_search (WikiPathways):

Input:
```
query
```
(string) - Search term
Output: WikiPathways search results
Use: Additional pathway context

STRING_functional_enrichment（主要富集工具）:

输入:
```
protein_ids
```
（基因符号数组）,
```
species
```
（整数，人类为9606）

输出:

{status: 'success', data: [{category, term, number_of_genes, number_of_genes_in_background, p_value, fdr, description, inputGenes, preferredNames}]}

用途: 针对GO、KEGG、Reactome、COMPARTMENTS、DISEASES进行全面富集分析

类别:

Process

（GO:BP）、

Function

（GO:MF）、

Component

（GO:CC）、

KEGG

、

Reactome

、

COMPARTMENTS

、

DISEASES

、

Keyword

、

PMID

注意: 这是主要的富集工具，一次调用可返回所有类别结果

ReactomeAnalysis_pathway_enrichment（Reactome特异性分析）:

输入:
```
identifiers
```
（字符串，空格分隔的基因符号，非数组）

输出:

{data: {token, pathways_found, pathways: [{pathway_id, name, p_value, fdr, entities_found, entities_total}]}}

用途: 进行带有层级结构的详细Reactome通路分析
注意: identifiers为空格分隔的字符串，而非数组

Reactome_map_uniprot_to_pathways（单个基因）:

输入:
```
id
```
（字符串）- UniProt登录号
输出: 通路对象的纯列表（无数据包装）
用途: 将单个蛋白质映射到Reactome通路

GO_get_annotations_for_gene（单个基因的GO注释）:

输入:
```
gene_id
```
（字符串）- 基因符号或ID
输出: GO注释对象的纯列表
用途: 获取单个基因的GO注释

kegg_search_pathway（KEGG通路搜索）:

输入:
```
query
```
（字符串）- 通路名称或关键词
输出: KEGG通路搜索结果
用途: 找到与空间发现相关的KEGG通路

WikiPathways_search（WikiPathways）:

输入:
```
query
```
（字符串）- 搜索词
输出: WikiPathways搜索结果
用途: 获取额外的通路背景

Workflow

工作流程

Global SVG enrichment: Run STRING_functional_enrichment on ALL SVGs
- Filter results by FDR < 0.05
- Separate by category (Process, Function, Component, KEGG, Reactome)
- Report top 10-15 per category
Reactome detailed analysis: Run ReactomeAnalysis_pathway_enrichment
- Report top pathways with FDR < 0.05
Per-domain enrichment (if spatial domains provided):
- Run STRING_functional_enrichment on each domain's gene set
- Compare enriched pathways across domains
- Identify domain-specific vs shared pathways
Compile pathway tables: Merge results from all enrichment tools

全局SVG富集: 对所有SVG运行
```
STRING_functional_enrichment
```
- 按FDR < 0.05过滤结果
- 按类别（Process、Function、Component、KEGG、Reactome）分类
- 每个类别报告排名前10-15的结果
Reactome详细分析: 运行
```
ReactomeAnalysis_pathway_enrichment
```
- 报告FDR < 0.05的顶级通路
各区域富集（若提供空间区域）:
- 对每个区域的基因集运行
```
STRING_functional_enrichment
```
- 比较各区域的富集通路
- 识别区域特异性通路与共享通路
整理通路表格: 合并所有富集工具的结果

Enrichment Interpretation

富集解读

Signaling pathways (RTK, Wnt, Notch, Hedgehog): Cell-cell communication
Metabolic pathways: Tissue metabolic zonation
Immune pathways: Immune infiltration/exclusion
ECM/adhesion pathways: Tissue structure and remodeling
Cell cycle/proliferation: Growth zones
Apoptosis/stress: Damage zones

信号通路（RTK、Wnt、Notch、Hedgehog）: 细胞间通讯
代谢通路: 组织代谢分区
免疫通路: 免疫浸润/排斥
ECM/黏附通路: 组织结构与重塑
细胞周期/增殖: 生长区域
凋亡/应激: 损伤区域

Phase 3: Spatial Domain Characterization

阶段3：空间区域表征

Objective: Characterize each spatial domain biologically and compare between domains.

目标: 从生物学角度表征每个空间区域，并进行区域间比较。

Tools Used

使用的工具

Uses the same tools as Phase 2 (STRING_functional_enrichment, ReactomeAnalysis) applied per-domain, plus:

HPA_get_biological_processes_by_gene (per-gene processes):

Input:
```
gene_name
```
(string)
Output: Biological processes associated with the gene
Use: Annotate domain marker genes

HPA_get_protein_interactions_by_gene (gene interactions):

Input:
```
gene_name
```
(string)
Output: Known protein interaction partners
Use: Build domain-specific interaction context

使用与阶段2相同的工具（

STRING_functional_enrichment

、

ReactomeAnalysis

）并应用于各区域，此外还有：

HPA_get_biological_processes_by_gene（单个基因的生物学过程）:

输入:
```
gene_name
```
（字符串）
输出: 与基因相关的生物学过程
用途: 注释区域标记基因

HPA_get_protein_interactions_by_gene（基因互作）:

输入:
```
gene_name
```
（字符串）
输出: 已知的蛋白质互作伙伴
用途: 构建区域特异性互作背景

Workflow

工作流程

For each spatial domain: a. Get marker gene list b. Run STRING_functional_enrichment on domain genes c. Identify top pathways, GO terms d. Assign likely cell type(s) based on marker genes:
- Epithelial: CDH1, EPCAM, KRT18, KRT19
- Mesenchymal/Fibroblast: VIM, COL1A1, COL3A1, FAP, ACTA2
- Immune T cell: CD3E, CD3D, CD4, CD8A, CD8B
- Immune B cell: CD19, CD20 (MS4A1), CD79A
- Macrophage: CD68, CD163, CSF1R
- Endothelial: PECAM1, VWF, CDH5
- Neuronal: SNAP25, SYP, MAP2, NEFL
- Hepatocyte: ALB, HNF4A, CYP3A4 e. Generate biological interpretation narrative
Compare domains:
- Differential pathways
- Unique vs shared genes
- Disease-relevant vs homeostatic regions
- Transition zones (shared genes between adjacent domains)

针对每个空间区域: a. 获取标记基因列表 b. 对区域基因集运行
```
STRING_functional_enrichment
```
c. 识别顶级通路、GO术语 d. 根据标记基因分配可能的细胞类型:
- 上皮细胞: CDH1、EPCAM、KRT18、KRT19
- 间充质/成纤维细胞: VIM、COL1A1、COL3A1、FAP、ACTA2
- 免疫T细胞: CD3E、CD3D、CD4、CD8A、CD8B
- 免疫B细胞: CD19、CD20（MS4A1）、CD79A
- 巨噬细胞: CD68、CD163、CSF1R
- 内皮细胞: PECAM1、VWF、CDH5
- 神经元: SNAP25、SYP、MAP2、NEFL
- 肝细胞: ALB、HNF4A、CYP3A4 e. 生成生物学解读描述
区域比较:
- 差异通路
- 独特基因与共享基因
- 疾病相关区域与稳态区域
- 过渡区域（相邻区域的共享基因）

Cell Type Assignment Rules

细胞类型分配规则

When user does not provide cell type annotations, infer from marker genes:

Check each gene against known cell type markers
Use HPA tissue/cell type expression data for validation
Report confidence level (high: 3+ markers match, medium: 2 markers, low: 1 marker)

当用户未提供细胞类型注释时，根据标记基因推断:

检查每个基因是否与已知细胞类型标记匹配
使用HPA组织/细胞类型表达数据进行验证
报告置信度等级（高: 3个及以上标记匹配，中: 2个标记匹配，低: 1个标记匹配）

Phase 4: Cell-Cell Interaction Inference

阶段4：细胞间互作推断

Objective: Predict cell-cell communication from spatial gene expression patterns.

目标: 从空间基因表达模式预测细胞间通讯。

Tools Used

使用的工具

STRING_get_interaction_partners (PPI network):

Input:
```
protein_ids
```
(array),
```
species
```
(int, 9606),
```
limit
```
(int),
```
confidence_score
```
(float, 0.7)

Output:

{status: 'success', data: [{preferredName_A, preferredName_B, score, nscore, fscore, pscore, ascore, escore, dscore, tscore}]}

Use: Find protein-protein interactions among SVGs
Score types: nscore=neighborhood, fscore=fusion, pscore=phylogenetic, ascore=coexpression, escore=experimental, dscore=database, tscore=textmining

STRING_get_protein_interactions (pairwise interactions):

Input:
```
protein_ids
```
(array),
```
species
```
(int, 9606)
Output: Interaction data between specified proteins
Use: Get interactions within a specific gene set

intact_search_interactions (IntAct database):

Input:
```
query
```
(string),
```
max
```
(int)
Output: Interaction data from IntAct
Use: Complement STRING with IntAct interactions

Reactome_get_interactor (Reactome interactions):

Input: Protein/gene identifier
Output: Reactome interaction data
Use: Pathway-level interaction context

DGIdb_get_drug_gene_interactions (drug-gene interactions):

Input:
```
genes
```
(array of strings)
Output: Drug-gene interaction data
Use: Identify druggable interaction nodes

STRING_get_interaction_partners（PPI网络）:

输入:
```
protein_ids
```
（数组）,
```
species
```
（整数，9606）,
```
limit
```
（整数）,
```
confidence_score
```
（浮点数，0.7）

输出:

{status: 'success', data: [{preferredName_A, preferredName_B, score, nscore, fscore, pscore, ascore, escore, dscore, tscore}]}

用途: 查找SVG之间的蛋白质-蛋白质互作
评分类型: nscore=邻域评分, fscore=融合评分, pscore=系统发育评分, ascore=共表达评分, escore=实验评分, dscore=数据库评分, tscore=文本挖掘评分

STRING_get_protein_interactions（成对互作）:

输入:
```
protein_ids
```
（数组）,
```
species
```
（整数，9606）
输出: 指定蛋白质之间的互作数据
用途: 获取特定基因集内的互作

intact_search_interactions（IntAct数据库）:

输入:
```
query
```
（字符串）,
```
max
```
（整数）
输出: IntAct数据库中的互作数据
用途: 用IntAct互作补充STRING结果

Reactome_get_interactor（Reactome互作）:

输入: 蛋白质/基因标识符
输出: Reactome互作数据
用途: 获取通路层面的互作背景

DGIdb_get_drug_gene_interactions（药物-基因互作）:

输入:
```
genes
```
（字符串数组）
输出: 药物-基因互作数据
用途: 识别可成药的互作节点

Ligand-Receptor Analysis

配体-受体分析

Known ligand-receptor pairs to check in SVG list:

Growth factors: EGF-EGFR, HGF-MET, VEGF-KDR, FGF-FGFR, PDGF-PDGFRA/B
Cytokines: TNF-TNFR, IL6-IL6R, IFNG-IFNGR, TGFB1-TGFBR1/2
Chemokines: CXCL12-CXCR4, CCL2-CCR2, CXCL10-CXCR3
Immune checkpoints: CD274(PD-L1)-PDCD1(PD-1), CD80/CD86-CTLA4, LGALS9-HAVCR2(TIM-3)
Notch signaling: DLL1/3/4-NOTCH1/2/3/4, JAG1/2-NOTCH1/2
Wnt signaling: WNT ligands-FZD receptors
Adhesion: CDH1-CDH1 (homotypic), ITGA/B integrins-ECM
Hedgehog: SHH-PTCH1

需要在SVG列表中检查的已知配体-受体对:

生长因子: EGF-EGFR、HGF-MET、VEGF-KDR、FGF-FGFR、PDGF-PDGFRA/B
细胞因子: TNF-TNFR、IL6-IL6R、IFNG-IFNGR、TGFB1-TGFBR1/2
趋化因子: CXCL12-CXCR4、CCL2-CCR2、CXCL10-CXCR3
免疫检查点: CD274(PD-L1)-PDCD1(PD-1)、CD80/CD86-CTLA4、LGALS9-HAVCR2(TIM-3)
Notch信号: DLL1/3/4-NOTCH1/2/3/4、JAG1/2-NOTCH1/2
Wnt信号: WNT配体-FZD受体
黏附: CDH1-CDH1（同型）、ITGA/B整合素-ECM
Hedgehog: SHH-PTCH1

Workflow

工作流程

Run STRING_get_interaction_partners on all SVGs
- Filter interactions with score > 0.7
- Identify hub genes (most connections)
Check for known ligand-receptor pairs in gene list
- Cross-reference with spatial domain assignments
- Identify potential cross-domain signaling
Build interaction network:
- Intra-domain interactions (within same spatial region)
- Inter-domain interactions (between different regions)
- Identify signaling axes (e.g., tumor-stroma, immune-tumor)
Map interactions to Reactome signaling pathways

对所有SVG运行
```
STRING_get_interaction_partners
```
- 过滤评分>0.7的互作
- 识别枢纽基因（连接数最多的基因）
检查基因列表中的已知配体-受体对
- 结合空间区域分配进行交叉参考
- 识别潜在的跨区域信号传导
构建互作网络:
- 区域内互作（同一空间区域内）
- 区域间互作（不同区域之间）
- 识别信号轴（如肿瘤-基质、免疫-肿瘤）
将互作映射到Reactome信号通路

Phase 5: Disease & Therapeutic Context

阶段5：疾病与治疗背景

Objective: Connect spatial findings to disease mechanisms and identify druggable targets.

目标: 将空间发现与疾病机制关联，识别可成药靶点。

Tools Used

使用的工具

OpenTargets_get_associated_targets_by_disease_efoId (disease genes):

Input:
```
efoId
```
(string),
```
size
```
(int)

Output:

{data: {disease: {associatedTargets: {count, rows: [{target: {id, approvedSymbol}, score}]}}}}

Use: Get disease-associated genes, overlap with SVGs

OpenTargets_get_target_tractability_by_ensemblID (druggability):

Input:
```
ensemblId
```
(string)
Output: Tractability data (small molecule, antibody, other modalities)
Use: Assess if spatial targets are druggable

OpenTargets_get_associated_drugs_by_target_ensemblID (drugs for target):

Input:
```
ensemblId
```
(string),
```
size
```
(int)
Output: Drug data for the target
Use: Find approved/clinical drugs targeting spatial genes

OpenTargets_get_drug_mechanisms_of_action_by_chemblId (drug mechanism):

Input:
```
chemblId
```
(string)
Output: Mechanism of action data
Use: Understand how drugs act on spatial targets

OpenTargets_target_disease_evidence (evidence linking target to disease):

Input:
```
ensemblId
```
(string),
```
efoId
```
(string)
Output: Evidence items linking target to disease
Use: Specific evidence for each spatial gene in disease

clinical_trials_search (clinical trials):

Input:

action

"search_studies"

condition

(string),

intervention

(string),

limit

(int)

Output:

{total_count, studies: [{nctId, title, status, conditions}]}

Use: Find clinical trials for spatial targets
NOTE:
```
action
```
MUST be
```
"search_studies"
```

DGIdb_get_gene_druggability (druggability categories):

Input:
```
genes
```
(array of strings)

Output:

{data: {genes: {nodes: [{name, geneCategories: [{name}]}]}}}

Use: Classify genes as druggable, kinase, GPCR, etc.

civic_search_genes (CIViC cancer evidence, if cancer):

Input: (no filter by name)
Output: Gene list from CIViC
Use: Check if SVGs have CIViC clinical evidence

OpenTargets_get_associated_targets_by_disease_efoId（疾病基因）:

输入:
```
efoId
```
（字符串）,
```
size
```
（整数）

输出:

{data: {disease: {associatedTargets: {count, rows: [{target: {id, approvedSymbol}, score}]}}}}

用途: 获取疾病相关基因，与SVG取交集

OpenTargets_get_target_tractability_by_ensemblID（成药性）:

输入:
```
ensemblId
```
（字符串）
输出: 成药性数据（小分子、抗体、其他模态）
用途: 评估空间靶点的成药性

OpenTargets_get_associated_drugs_by_target_ensemblID（靶点相关药物）:

输入:
```
ensemblId
```
（字符串）,
```
size
```
（整数）
输出: 靶点相关药物数据
用途: 找到针对空间基因的已获批/临床阶段药物

OpenTargets_get_drug_mechanisms_of_action_by_chemblId（药物机制）:

输入:
```
chemblId
```
（字符串）
输出: 作用机制数据
用途: 理解药物作用于空间靶点的机制

OpenTargets_target_disease_evidence（靶点与疾病关联的证据）:

输入:
```
ensemblId
```
（字符串）,
```
efoId
```
（字符串）
输出: 连接靶点与疾病的证据条目
用途: 获取每个空间基因在疾病中的具体证据

clinical_trials_search（临床试验）:

输入:
```
action
```
=
```
"search_studies"
```
,
```
condition
```
（字符串）,
```
intervention
```
（字符串）,
```
limit
```
（整数）

输出:

{total_count, studies: [{nctId, title, status, conditions}]}

用途: 查找针对空间靶点的临床试验
注意:
```
action
```
必须设为
```
"search_studies"
```

DGIdb_get_gene_druggability（成药性分类）:

输入:
```
genes
```
（字符串数组）

输出:

{data: {genes: {nodes: [{name, geneCategories: [{name}]}]}}}

用途: 将基因分类为可成药、激酶、GPCR等类型

civic_search_genes（CIViC癌症证据，若为癌症）:

输入: （无名称过滤）
输出: CIViC中的基因列表
用途: 检查SVG是否有CIViC临床证据

Workflow

工作流程

Disease gene overlap (if disease context provided): a. Get disease-associated targets from OpenTargets b. Intersect with SVGs c. For overlapping genes, get specific evidence
Druggable target identification: a. Run DGIdb_get_gene_druggability on all SVGs b. For druggable genes, check OpenTargets tractability c. Get approved drugs for druggable spatial targets
Clinical trials: a. Search for trials targeting spatial genes in the disease context b. Prioritize trials for genes in disease-enriched spatial domains
Cancer-specific (if cancer): a. Check CIViC for clinical evidence b. Get mutation prevalence from cBioPortal (if specific mutations known) c. Check immune checkpoint genes in spatial data

疾病基因交集（若提供疾病背景）: a. 从OpenTargets获取疾病相关靶点 b. 与SVG取交集 c. 对交集基因获取具体证据
可成药靶点识别: a. 对所有SVG运行
```
DGIdb_get_gene_druggability
```
b. 对可成药基因，检查OpenTargets成药性 c. 获取针对可成药空间靶点的已获批药物
临床试验: a. 在疾病背景下搜索针对空间基因的试验 b. 优先关注疾病富集区域中基因的试验
癌症特异性分析（若为癌症）: a. 检查CIViC中的临床证据 b. 从cBioPortal获取突变频率（若已知特定突变） c. 检查空间数据中的免疫检查点基因

Phase 6: Multi-Modal Integration

阶段6：多模态整合

Objective: Integrate protein, RNA, and metabolite spatial data when available.

目标: 若有可用数据，整合蛋白质、RNA和代谢物空间数据。

Tools Used

使用的工具

HPA_get_subcellular_location (protein localization):

Input:
```
gene_name
```
(string)

Output:

{gene_name, main_locations, additional_locations, location_summary}

Use: Compare mRNA spatial pattern with protein subcellular location

HPA_get_rna_expression_in_specific_tissues (tissue RNA):

Input:
```
ensembl_id
```
(string),
```
tissue_name
```
(string)
Output: Expression data for specific tissue
Use: Validate spatial expression against bulk tissue data

Reactome_map_uniprot_to_pathways (metabolic pathways):

Input:
```
id
```
(string) - UniProt accession
Output: List of pathways
Use: Map genes to metabolic pathways for metabolomics integration

kegg_get_pathway_info (KEGG pathway details):

Input:
```
pathway_id
```
(string) - KEGG pathway ID
Output: Pathway information including metabolites
Use: Link spatial genes to metabolic pathways and metabolites

HPA_get_subcellular_location（蛋白质定位）:

输入:
```
gene_name
```
（字符串）

输出:

{gene_name, main_locations, additional_locations, location_summary}

用途: 比较mRNA空间模式与蛋白质亚细胞定位

HPA_get_rna_expression_in_specific_tissues（组织RNA表达）:

输入:
```
ensembl_id
```
（字符串）,
```
tissue_name
```
（字符串）
输出: 特定组织的表达数据
用途: 验证空间表达与 bulk 组织数据的一致性

Reactome_map_uniprot_to_pathways（代谢通路）:

输入:
```
id
```
（字符串）- UniProt登录号
输出: 通路列表
用途: 将基因映射到代谢通路以进行代谢组整合

kegg_get_pathway_info（KEGG通路详情）:

输入:
```
pathway_id
```
（字符串）- KEGG通路ID
输出: 包含代谢物的通路信息
用途: 将空间基因与代谢通路和代谢物关联

Workflow

工作流程

RNA-Protein concordance (if protein data provided): a. For each gene with both RNA and protein data:
- Compare spatial RNA pattern with protein detection
- Check HPA for known post-transcriptional regulation
- Note concordant (expected) vs discordant (interesting) patterns
Subcellular context: a. Map spatial RNA localization to protein subcellular location (HPA) b. Secreted proteins -> likely paracrine signaling c. Membrane proteins -> cell surface markers d. Nuclear proteins -> transcription factors
Metabolic integration (if metabolomics available): a. Map genes to metabolic pathways (Reactome, KEGG) b. Link detected metabolites to enzyme-encoding genes c. Identify spatial metabolic heterogeneity d. Check for known metabolic zonation patterns

RNA-蛋白质一致性（若有蛋白质数据）: a. 对同时有RNA和蛋白质数据的每个基因:
- 比较RNA空间模式与蛋白质检测结果
- 检查HPA中已知的转录后调控
- 记录一致（预期）与不一致（值得关注）的模式
亚细胞背景: a. 将RNA空间定位与蛋白质亚细胞定位（HPA）关联 b. 分泌蛋白 -> 可能为旁分泌信号 c. 膜蛋白 -> 细胞表面标记 d. 核蛋白 -> 转录因子
代谢整合（若有代谢组数据）: a. 将基因映射到代谢通路（Reactome、KEGG） b. 将检测到的代谢物与编码酶的基因关联 c. 识别空间代谢异质性 d. 检查已知的代谢分区模式

Phase 7: Immune Microenvironment (Cancer/Inflammation)

阶段7：免疫微环境（癌症/炎症）

Objective: Characterize immune cell composition and checkpoint expression in spatial context.

目标: 在空间背景下表征免疫细胞组成与检查点表达。

Conditions for Activation

激活条件

Only execute if:

Disease context is cancer, autoimmune, or inflammatory
SVGs include immune markers (CD3E, CD8A, CD68, CD163, etc.)
User specifically asks about immune patterns

仅在以下情况执行:

疾病背景为癌症、自身免疫病或炎症
SVG包含免疫标记（CD3E、CD8A、CD68、CD163等）
用户专门询问免疫模式

Tools Used

使用的工具

STRING_functional_enrichment (immune pathway enrichment):

Applied to immune-relevant SVGs
Filter for immune-related GO terms and pathways

OpenTargets_get_target_tractability_by_ensemblID (checkpoint druggability):

Applied to immune checkpoint genes
Check for approved immunotherapies

iedb_search_epitopes (epitope data):

Input:
```
organism_name
```
(string),
```
source_antigen_name
```
(string)
Output:
```
{status, data, count}
```
Use: Check if spatial antigens have known epitopes

STRING_functional_enrichment（免疫通路富集）:

应用于免疫相关SVG
过滤免疫相关GO术语与通路

OpenTargets_get_target_tractability_by_ensemblID（检查点成药性）:

应用于免疫检查点基因
检查已获批的免疫疗法

iedb_search_epitopes（表位数据）:

输入:
```
organism_name
```
（字符串）,
```
source_antigen_name
```
（字符串）
输出:
```
{status, data, count}
```
用途: 检查空间抗原是否有已知表位

Immune Cell Markers Reference

免疫细胞标记参考

Cell Type	Key Markers	Extended Markers
CD8+ T cell	CD8A, CD8B	GZMA, GZMB, PRF1, IFNG
CD4+ T cell	CD4	IL2, IL4, IL17A, FOXP3 (Treg)
Regulatory T cell	FOXP3, IL2RA	CTLA4, TIGIT
B cell	CD19, MS4A1, CD79A	IGHG1, IGHM
Plasma cell	SDC1 (CD138), XBP1	IGHG1, MZB1
M1 Macrophage	CD68, NOS2, TNF	IL1B, CXCL10
M2 Macrophage	CD68, CD163, MRC1	ARG1, IL10
Dendritic cell	ITGAX (CD11c), HLA-DRA	CD80, CD86
NK cell	NCAM1 (CD56), NKG7	GNLY, KLRD1
Neutrophil	FCGR3B, CXCR2	S100A8, S100A9
Mast cell	KIT, TPSAB1	CPA3, HDC

细胞类型	关键标记	扩展标记
CD8+ T细胞	CD8A、CD8B	GZMA、GZMB、PRF1、IFNG
CD4+ T细胞	CD4	IL2、IL4、IL17A、FOXP3（Treg）
调节性T细胞	FOXP3、IL2RA	CTLA4、TIGIT
B细胞	CD19、MS4A1、CD79A	IGHG1、IGHM
浆细胞	SDC1（CD138）、XBP1	IGHG1、MZB1
M1巨噬细胞	CD68、NOS2、TNF	IL1B、CXCL10
M2巨噬细胞	CD68、CD163、MRC1	ARG1、IL10
树突状细胞	ITGAX（CD11c）、HLA-DRA	CD80、CD86
NK细胞	NCAM1（CD56）、NKG7	GNLY、KLRD1
中性粒细胞	FCGR3B、CXCR2	S100A8、S100A9
肥大细胞	KIT、TPSAB1	CPA3、HDC

Immune Checkpoint Reference

免疫检查点参考

Checkpoint	Gene	Ligand	Therapeutic Antibody
PD-1/PD-L1	PDCD1/CD274	CD274, PDCD1LG2	Pembrolizumab, Nivolumab, Atezolizumab
CTLA-4	CTLA4	CD80, CD86	Ipilimumab
TIM-3	HAVCR2	LGALS9	Sabatolimab
LAG-3	LAG3	HLA class II	Relatlimab
TIGIT	TIGIT	PVR, PVRL2	Tiragolumab
VISTA	VSIR	PSGL1	-

检查点	基因	配体	治疗性抗体
PD-1/PD-L1	PDCD1/CD274	CD274、PDCD1LG2	Pembrolizumab、Nivolumab、Atezolizumab
CTLA-4	CTLA4	CD80、CD86	Ipilimumab
TIM-3	HAVCR2	LGALS9	Sabatolimab
LAG-3	LAG3	HLA II类分子	Relatlimab
TIGIT	TIGIT	PVR、PVRL2	Tiragolumab
VISTA	VSIR	PSGL1	-

Workflow

工作流程

Identify immune-related SVGs from marker reference
Classify immune cell types present per spatial domain
Check immune checkpoint expression
Assess immune infiltration patterns:
- Hot (T cell infiltrated) vs Cold (immune desert) vs Excluded
Identify potential immunotherapy targets
Check for tertiary lymphoid structures (B cell + T cell clusters)

从标记参考中识别免疫相关SVG
分类每个空间区域中存在的免疫细胞类型
检查免疫检查点表达
评估免疫浸润模式:
- 热区（T细胞浸润）、冷区（免疫荒漠）、排斥区
识别潜在的免疫治疗靶点
检查三级淋巴结构（B细胞+T细胞簇）

Phase 8: Literature & Validation Context

阶段8：文献与验证背景

Objective: Provide literature evidence for spatial findings and suggest validation experiments.

目标: 为空间发现提供文献证据，建议实验验证方法。

Tools Used

使用的工具

PubMed_search_articles (literature search):

Input:
```
query
```
(string),
```
max_results
```
(int)

Output: List of

[{pmid, title, authors, journal, pub_date, doi}]

Use: Find published evidence for spatial patterns

openalex_literature_search (broader literature):

Input:
```
query
```
(string),
```
per_page
```
(int)
Output: List of works with titles, DOIs, abstracts
Use: Complement PubMed with preprints and broader coverage

PubMed_search_articles（文献检索）:

输入:
```
query
```
（字符串）,
```
max_results
```
（整数）

输出:

[{pmid, title, authors, journal, pub_date, doi}]

列表

用途: 查找空间模式的已发表证据

openalex_literature_search（更广泛的文献）:

输入:
```
query
```
（字符串）,
```
per_page
```
（整数）
输出: 包含标题、DOI、摘要的文献列表
用途: 用预印本和更广泛的覆盖范围补充PubMed

Literature Search Strategy

文献检索策略

Tissue + spatial:
```
"{tissue} spatial transcriptomics"
```
- e.g., "liver spatial transcriptomics"
Disease + spatial:
```
"{disease} spatial omics"
```
- e.g., "breast cancer spatial transcriptomics"
Gene + tissue:
```
"{top_gene} {tissue} expression"
```
for key SVGs
Zonation (if relevant):
```
"{tissue} zonation gene expression"
```
Technology:
```
"{technology} {tissue}"
```
- e.g., "Visium breast cancer"

组织+空间:
```
"{tissue} spatial transcriptomics"
```
- 例如"liver spatial transcriptomics"
疾病+空间:
```
"{disease} spatial omics"
```
- 例如"breast cancer spatial transcriptomics"
基因+组织: 针对关键SVG使用
```
"{top_gene} {tissue} expression"
```
分区（如适用）:
```
"{tissue} zonation gene expression"
```
技术:
```
"{technology} {tissue}"
```
- 例如"Visium breast cancer"

Validation Recommendations Template

验证建议模板

Priority	Target	Method	Rationale	Feasibility
High	Key SVG	smFISH / RNAscope	Validate spatial pattern at single-molecule level	Medium
High	Druggable target	IHC on serial sections	Confirm protein expression in spatial domain	High
High	Ligand-receptor pair	Proximity ligation assay (PLA)	Confirm physical interaction at tissue level	Medium
Medium	Domain markers	Multiplexed IF (CODEX/IBEX)	Validate multiple markers simultaneously	Low-Medium
Medium	Pathway	Spatial metabolomics (MALDI/DESI)	Confirm metabolic pathway activity	Low
Low	Novel interaction	Co-culture + conditioned media	Functional validation of predicted interaction	Medium

优先级	靶点	方法	理由	可行性
高	关键SVG	smFISH / RNAscope	在单分子水平验证空间模式	中等
高	可成药靶点	连续切片IHC	确认蛋白质在空间区域中的表达	高
高	配体-受体对	邻近连接实验（PLA）	在组织水平确认物理互作	中等
中	区域标记	多重免疫荧光（CODEX/IBEX）	同时验证多个标记	低-中等
中	通路	空间代谢组学（MALDI/DESI）	确认代谢通路活性	低
低	新型互作	共培养+条件培养基	功能验证预测的互作	中等

Workflow

工作流程

Search PubMed for tissue + disease + spatial transcriptomics
Search for known spatial patterns in the tissue type
Cross-reference findings with published spatial atlas data
Generate validation recommendations based on:
- Novelty of finding (novel patterns need more validation)
- Clinical relevance (druggable targets prioritized)
- Technical feasibility
Cite relevant methodology papers for each validation approach

检索组织+疾病+空间转录组相关的PubMed文献
检索该组织类型的已知空间模式
将发现与已发表的空间图谱数据交叉参考
根据以下因素生成验证建议:
- 发现的新颖性（新型模式需要更多验证）
- 临床相关性（可成药靶点优先）
- 技术可行性
为每种验证方法引用相关的方法学文献

Tool Parameter Reference (CRITICAL)

工具参数参考（至关重要）

Verified Parameter Names

已验证的参数名称

Tool	Parameter	CORRECT	Common MISTAKE	Notes
`MyGene_query_genes`	query	`query`	`q`	Filter results by `symbol` field
`STRING_functional_enrichment`	identifiers	`protein_ids` (array)	`identifiers`	Also needs `species=9606`
`STRING_get_interaction_partners`	identifiers	`protein_ids` (array)	`identifiers`	`limit` , `confidence_score` optional
`ReactomeAnalysis_pathway_enrichment`	genes	`identifiers` (string)	Array	SPACE-SEPARATED string, NOT array
`HPA_get_subcellular_location`	gene	`gene_name`	`ensembl_id`	Uses gene symbol
`HPA_get_cancer_prognostics_by_gene`	gene	`ensembl_id`	`gene_name`	Uses Ensembl ID, NOT symbol
`HPA_get_rna_expression_by_source`	params	`gene_name` , `source_type` , `source_name`	-	ALL 3 required
`HPA_get_rna_expression_in_specific_tissues`	gene	`ensembl_id`	`gene_name`	Uses Ensembl ID
`OpenTargets_get_target_tractability_by_ensemblID`	target	`ensemblId`	`ensemblID`	camelCase
`OpenTargets_get_associated_drugs_by_target_ensemblID`	target	`ensemblId` , `size`	-	Both REQUIRED
`OpenTargets_get_associated_targets_by_disease_efoId`	disease	`efoId`	`diseaseId`	Returns {data: {disease: {associatedTargets}}}
`DGIdb_get_gene_druggability`	genes	`genes` (array)	`gene_name`	Array of strings
`DGIdb_get_drug_gene_interactions`	genes	`genes` (array)	`gene_name`	Array of strings
`clinical_trials_search`	action	`action='search_studies'`	Missing action	`action` is REQUIRED
`ensembl_lookup_gene`	species	`species='homo_sapiens'`	No species	REQUIRED parameter
GTEx tools	operation	`operation` (SOAP)	Missing	All GTEx tools need `operation` parameter
`HPA_get_comprehensive_gene_details_by_ensembl_id`	all params	ALL 5 required: `ensembl_id` , `include_isoforms` , `include_images` , `include_antibodies` , `include_expression`	Missing booleans	Set booleans to False except expression
GTEx tools	gencode	`gencode_id` (array)	`gene_id`	Requires versioned GENCODE ID

工具	参数	正确名称	常见错误	说明
`MyGene_query_genes`	query	`query`	`q`	按 `symbol` 字段过滤结果
`STRING_functional_enrichment`	identifiers	`protein_ids` （数组）	`identifiers`	还需要 `species=9606`
`STRING_get_interaction_partners`	identifiers	`protein_ids` （数组）	`identifiers`	`limit` 、 `confidence_score` 为可选参数
`ReactomeAnalysis_pathway_enrichment`	genes	`identifiers` （字符串）	数组	空格分隔的字符串，而非数组
`HPA_get_subcellular_location`	gene	`gene_name`	`ensembl_id`	使用基因符号
`HPA_get_cancer_prognostics_by_gene`	gene	`ensembl_id`	`gene_name`	使用Ensembl ID，而非符号
`HPA_get_rna_expression_by_source`	params	`gene_name` , `source_type` , `source_name`	-	所有3个参数必填
`HPA_get_rna_expression_in_specific_tissues`	gene	`ensembl_id`	`gene_name`	使用Ensembl ID
`OpenTargets_get_target_tractability_by_ensemblID`	target	`ensemblId`	`ensemblID`	小驼峰命名
`OpenTargets_get_associated_drugs_by_target_ensemblID`	target	`ensemblId` , `size`	-	两者均为必填项
`OpenTargets_get_associated_targets_by_disease_efoId`	disease	`efoId`	`diseaseId`	返回 `{data: {disease: {associatedTargets}}}`
`DGIdb_get_gene_druggability`	genes	`genes` （数组）	`gene_name`	字符串数组
`DGIdb_get_drug_gene_interactions`	genes	`genes` （数组）	`gene_name`	字符串数组
`clinical_trials_search`	action	`action='search_studies'`	缺少action	`action` 为必填项
`ensembl_lookup_gene`	species	`species='homo_sapiens'`	无species参数	必填参数
GTEx工具	operation	`operation` （SOAP）	缺少	所有GTEx工具都需要 `operation` 参数
`HPA_get_comprehensive_gene_details_by_ensembl_id`	所有参数	5个参数均必填: `ensembl_id` , `include_isoforms` , `include_images` , `include_antibodies` , `include_expression`	缺少布尔值参数	除expression外，其他布尔值设为False
GTEx工具	gencode	`gencode_id` （数组）	`gene_id`	需要带版本的GENCODE ID

Response Format Reference

响应格式参考

Tool	Response Format	Key Fields
`STRING_functional_enrichment`	`{status, data: [{category, term, description, p_value, fdr, inputGenes}]}`	Filter by FDR < 0.05
`ReactomeAnalysis_pathway_enrichment`	`{data: {pathways: [{pathway_id, name, p_value, fdr, entities_found, entities_total}]}}`	Top 20 returned
`STRING_get_interaction_partners`	`{status, data: [{preferredName_A, preferredName_B, score}]}`	Score > 0.7 for high confidence
`MyGene_query_genes`	`{hits: [{_id, symbol, name, ensembl: {gene}, entrezgene}]}`	Filter by exact symbol match
`HPA_get_subcellular_location`	`{gene_name, main_locations: [], additional_locations: [], location_summary}`	Direct dict response
`OpenTargets_get_target_tractability_by_ensemblID`	`{data: {target: {id, tractability: [{label, modality, value}]}}}`	Check value=true
`DGIdb_get_gene_druggability`	`{data: {genes: {nodes: [{name, geneCategories: [{name}]}]}}}`	GraphQL response
`PubMed_search_articles`	Plain list of `[{pmid, title, authors, journal, pub_date}]`	No data wrapper
`clinical_trials_search`	`{total_count, studies: [{nctId, title, status, conditions}]}`	total_count can be None

工具	响应格式	关键字段
`STRING_functional_enrichment`	`{status, data: [{category, term, description, p_value, fdr, inputGenes}]}`	按FDR < 0.05过滤
`ReactomeAnalysis_pathway_enrichment`	`{data: {pathways: [{pathway_id, name, p_value, fdr, entities_found, entities_total}]}}`	返回排名前20的结果
`STRING_get_interaction_partners`	`{status, data: [{preferredName_A, preferredName_B, score}]}`	评分>0.7为高置信度
`MyGene_query_genes`	`{hits: [{_id, symbol, name, ensembl: {gene}, entrezgene}]}`	按精确符号匹配过滤
`HPA_get_subcellular_location`	`{gene_name, main_locations: [], additional_locations: [], location_summary}`	直接字典响应
`OpenTargets_get_target_tractability_by_ensemblID`	`{data: {target: {id, tractability: [{label, modality, value}]}}}`	检查value=true
`DGIdb_get_gene_druggability`	`{data: {genes: {nodes: [{name, geneCategories: [{name}]}]}}}`	GraphQL响应
`PubMed_search_articles`	`[{pmid, title, authors, journal, pub_date}]` 纯列表	无数据包装
`clinical_trials_search`	`{total_count, studies: [{nctId, title, status, conditions}]}`	total_count可为空

Fallback Strategies

fallback策略

Pathway Enrichment

通路富集

Primary: STRING_functional_enrichment (most comprehensive, one call)
Fallback: ReactomeAnalysis_pathway_enrichment (Reactome-specific)
Default: Individual gene GO annotations (GO_get_annotations_for_gene)

首选:
```
STRING_functional_enrichment
```
（最全面，一次调用）
备选:
```
ReactomeAnalysis_pathway_enrichment
```
（Reactome特异性）
默认: 单个基因GO注释（
```
GO_get_annotations_for_gene
```
）

Tissue Expression

组织表达

Primary: HPA_get_rna_expression_by_source
Fallback: HPA_get_comprehensive_gene_details_by_ensembl_id
Default: Note "tissue expression data unavailable"

首选:
```
HPA_get_rna_expression_by_source
```

备选:

HPA_get_comprehensive_gene_details_by_ensembl_id

默认: 标注“组织表达数据不可用”

Disease Association

疾病关联

Primary: OpenTargets_get_associated_targets_by_disease_efoId
Fallback: OpenTargets_target_disease_evidence (per gene)
Default: Skip disease section if no disease context

首选:

OpenTargets_get_associated_targets_by_disease_efoId

备选:
```
OpenTargets_target_disease_evidence
```
（每个基因）
默认: 若无疾病背景，跳过疾病章节

Drug Information

药物信息

Primary: OpenTargets_get_associated_drugs_by_target_ensemblID
Fallback: DGIdb_get_drug_gene_interactions
Default: Note "no approved drugs identified"

首选:

OpenTargets_get_associated_drugs_by_target_ensemblID

备选:
```
DGIdb_get_drug_gene_interactions
```
默认: 标注“未识别到已获批药物”

Literature

文献

Primary: PubMed_search_articles
Fallback: openalex_literature_search
Default: Note "no spatial-specific literature found"

首选:
```
PubMed_search_articles
```
备选:
```
openalex_literature_search
```
默认: 标注“未找到空间特异性文献”

Common Use Cases

常见用例

Use Case 1: Cancer Spatial Heterogeneity

用例1：癌症空间异质性

Input: Visium data from breast cancer with 5 spatial domains (tumor core, tumor margin, stroma, immune infiltrate, normal tissue) and 200 SVGs.

Analysis focus:

Tumor-specific pathways (proliferation, DNA repair)
Immune infiltration patterns (hot vs cold)
Tumor-stroma interactions (CAF signaling)
Druggable targets in tumor core
Immune checkpoint expression patterns
Prognostic genes per domain

输入: 乳腺癌Visium数据，包含5个空间区域（肿瘤核心、肿瘤边缘、基质、免疫浸润区、正常组织）和200个SVG。

分析重点:

肿瘤特异性通路（增殖、DNA修复）
免疫浸润模式（热区vs冷区）
肿瘤-基质互作（CAF信号）
肿瘤核心区的可成药靶点
免疫检查点表达模式
各区域的预后基因

Use Case 2: Brain Tissue Zonation

用例2：脑组织分区

Input: MERFISH data from hippocampus with cell-type specific genes and neuronal subtype markers.

Analysis focus:

Neuronal subtype characterization
Synaptic signaling pathways
Neurotransmitter receptor distribution
Known hippocampal zonation patterns (CA1, CA3, DG)
Neurodegenerative disease gene overlap

输入: 海马体MERFISH数据，包含细胞类型特异性基因和神经元亚型标记。

分析重点:

神经元亚型表征
突触信号通路
神经递质受体分布
已知海马体分区模式（CA1、CA3、DG）
神经退行性疾病基因交集

Use Case 3: Liver Metabolic Zonation

用例3：肝脏代谢分区

Input: Spatial transcriptomics of liver with periportal vs pericentral gene gradients.

Analysis focus:

Metabolic enzyme distribution (CYP450, gluconeogenesis, lipogenesis)
Wnt signaling gradient (known zonation regulator)
Oxygen gradient-responsive genes
Drug metabolism enzyme spatial patterns
Liver disease gene overlap

输入: 肝脏空间转录组数据，包含门脉周与中央静脉周基因梯度。

分析重点:

代谢酶分布（CYP450、糖异生、脂肪生成）
Wnt信号梯度（已知分区调控因子）
氧梯度响应基因
药物代谢酶空间模式
肝脏疾病基因交集

Use Case 4: Tumor-Immune Interface

用例4：肿瘤-免疫界面

Input: DBiTplus data from melanoma with spatial protein + RNA data showing tumor-immune boundary.

Analysis focus:

Immune cell composition at boundary
Checkpoint ligand-receptor pairs
Immune exclusion mechanisms
Immunotherapy target identification
Multi-modal (RNA + protein) concordance

输入: 黑色素瘤DBiTplus数据，包含显示肿瘤-免疫边界的空间蛋白质+RNA数据。

分析重点:

边界处的免疫细胞组成
检查点配体-受体对
免疫排斥机制
免疫治疗靶点识别
多模态（RNA+蛋白质）一致性

Use Case 5: Developmental Spatial Patterns

用例5：发育空间模式

Input: Spatial transcriptomics of embryonic tissue with developmental patterning genes.

Analysis focus:

Morphogen gradients (Wnt, BMP, FGF, SHH)
Transcription factor spatial patterns
Cell fate determination genes
Developmental signaling pathways
Comparison to adult tissue patterns

输入: 胚胎组织空间转录组数据，包含发育模式基因。

分析重点:

形态发生素梯度（Wnt、BMP、FGF、SHH）
转录因子空间模式
细胞命运决定基因
发育信号通路
与成年组织模式的比较

Use Case 6: Disease Progression Mapping

用例6：疾病进展图谱

Input: Spatial data from neurodegenerative tissue showing disease gradient from affected to unaffected regions.

Analysis focus:

Disease gene expression gradient
Inflammatory response spatial pattern
Neuronal loss markers
Glial activation patterns
Therapeutic window identification

输入: 神经退行性组织空间数据，显示从受影响区域到未受影响区域的疾病梯度。

分析重点:

疾病基因表达梯度
炎症反应空间模式
神经元丢失标记
胶质细胞激活模式
治疗窗口识别

Limitations & Known Issues

局限性与已知问题

Database-Specific

数据库特异性

Enrichment:
```
enrichr_gene_enrichment_analysis
```
returns connectivity graph (107MB), NOT standard enrichment. Use
```
STRING_functional_enrichment
```
instead
GTEx: SOAP-style tools requiring
```
operation
```
parameter; needs versioned GENCODE IDs (e.g.,
```
ENSG00000141510.16
```
)
HPA: Some tools use
```
gene_name
```
, others use
```
ensembl_id
```
- check parameter reference
OpenTargets: Disease IDs use underscore format (
```
MONDO_0007254
```
), not colon
cBioPortal_get_cancer_studies: BROKEN - has literal
```
{limit}
```
in URL causing 400 error

富集:
```
enrichr_gene_enrichment_analysis
```
返回连接图（107MB），而非标准富集结果。请使用
```
STRING_functional_enrichment
```
替代
GTEx: 需要
```
operation
```
参数的SOAP风格工具；需要带版本的GENCODE ID（例如
```
ENSG00000141510.16
```
）
HPA: 部分工具使用
```
gene_name
```
，其他使用
```
ensembl_id
```
- 请查阅参数参考
OpenTargets: 疾病ID使用下划线格式（
```
MONDO_0007254
```
），而非冒号
cBioPortal_get_cancer_studies: 已损坏 - URL中包含字面量
```
{limit}
```
导致400错误

Conceptual

概念性

No raw spatial data processing: This skill analyzes gene LISTS, not raw spatial matrices (Seurat/Scanpy/squidpy handle raw data)
No spatial statistics: Cannot perform Moran's I, spatial autocorrelation, or variogram analysis
No image analysis: Cannot process H&E or fluorescence images
No deconvolution: Cannot perform cell type deconvolution (use BayesSpace, cell2location, RCTD externally)
Ligand-receptor inference: Based on gene co-expression + known pairs, not spatial proximity statistics (use CellChat, NicheNet, COMMOT externally)

无原始空间数据处理: 该技能分析基因列表，而非原始空间矩阵（Seurat/Scanpy/squidpy处理原始数据）
无空间统计: 无法执行Moran's I、空间自相关或变异函数分析
无图像分析: 无法处理H&E或荧光图像
无去卷积: 无法执行细胞类型去卷积（请外部使用BayesSpace、cell2location、RCTD）
配体-受体推断: 基于基因共表达+已知对，而非空间邻近统计（请外部使用CellChat、NicheNet、COMMOT）

Technical

技术性

Large gene lists: >200 genes may slow STRING queries; batch or sample
Response format variability: Always check both dict and list response types
Rate limits: STRING and OpenTargets may throttle frequent requests

大基因列表: >200个基因可能减慢STRING查询；请批量处理或抽样
响应格式多变: 始终检查字典和列表两种响应类型
速率限制: STRING和OpenTargets可能对频繁请求进行限流

Summary

总结

Spatial Multi-Omics Analysis skill provides:

Gene characterization (ID resolution, function, localization, tissue expression)
Pathway & functional enrichment (STRING, Reactome, GO, KEGG)
Spatial domain characterization (per-domain and cross-domain comparison)
Cell-cell interaction inference (PPI, ligand-receptor, signaling pathways)
Disease & therapeutic context (disease genes, druggable targets, clinical trials)
Multi-modal integration (RNA-protein concordance, metabolic pathways)
Immune microenvironment characterization (cell types, checkpoints, immunotherapy)
Literature context & validation recommendations

Outputs: Comprehensive markdown report with Spatial Omics Integration Score (0-100) Best for: Biological interpretation of spatial omics experiments (post-processing after spatial data analysis tools) Uses: 70+ ToolUniverse tools across 9 analysis phases Time: ~10-20 minutes depending on gene list size and analysis scope

空间多组学分析技能提供:

基因表征（ID解析、功能、定位、组织表达）
通路与功能富集（STRING、Reactome、GO、KEGG）
空间区域表征（单区域与跨区域比较）
细胞间互作推断（PPI、配体-受体、信号通路）
疾病与治疗背景（疾病基因、可成药靶点、临床试验）
多模态整合（RNA-蛋白质一致性、代谢通路）
免疫微环境表征（细胞类型、检查点、免疫治疗）
文献背景与验证建议

输出: 包含空间多组学整合评分（0-100）的全面Markdown报告 最佳适用场景: 空间多组学实验的生物学解读（空间数据分析工具后处理） 使用工具: 9个分析阶段中使用70余种ToolUniverse工具耗时: 约10-20分钟，取决于基因列表大小与分析范围