AgentSkillsCN

literature-search

分析来自 GEO 数据集的 DNA 微阵列数据。适用于被要求分析微阵列数据、进行差异表达分析,或在获得 GEO 访问号(GSE*)时使用。支持 Affymetrix(CEL 文件)与 Illumina(BeadChip)平台,提供 QC、标准化、统计建模以及基因 ID 映射功能。

SKILL.md
--- frontmatter
name: literature-search
description: Search PubMed for scientific literature. Use when asked to find papers, search literature, look up publications, or gather references on a topic. Returns article titles, authors, abstracts, and PMIDs.
user-invocable: true

PubMed Literature Search Skill

This skill searches PubMed using the NCBI E-utilities API to find relevant scientific literature.

How to Use

When the user asks to search for literature, use the WebFetch tool to query PubMed's E-utilities API.

Step 1: Search for Articles

Use ESearch to find PMIDs matching a query:

code
https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=pubmed&term=<QUERY>&retmode=json&retmax=<COUNT>

Parameters:

  • term: URL-encoded search query (use + for spaces, standard PubMed syntax)
  • retmax: Maximum results to return (default 20, max 10000)
  • retstart: Offset for pagination
  • sort: Sort order (relevance, pub_date, first_author)

Example query for vitiligo transcriptomics:

code
https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=pubmed&term=vitiligo+transcriptome&retmode=json&retmax=20

Step 2: Get Article Details

Use ESummary to retrieve article metadata from PMIDs:

code
https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi?db=pubmed&id=<PMID_LIST>&retmode=json

Parameters:

  • id: Comma-separated list of PMIDs from Step 1

Step 3: Get Abstracts (Optional)

Use EFetch to retrieve full abstracts:

code
https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=pubmed&id=<PMID_LIST>&rettype=abstract&retmode=text

Search Syntax

PubMed supports advanced search operators:

OperatorExampleDescription
ANDvitiligo AND transcriptomeBoth terms required
ORvitiligo OR depigmentationEither term
NOTvitiligo NOT mouseExclude term
[tiab]vitiligo[tiab]Title/abstract only
[au]smith[au]Author search
[jour]nature[jour]Journal search
[pdat]2020:2024[pdat]Publication date range
[mesh]vitiligo[mesh]MeSH term
"phrase""gene expression"Exact phrase

Example Workflow

When user asks: "Find recent papers on vitiligo RNA-seq"

  1. Search PubMed:

    code
    WebFetch: https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=pubmed&term=vitiligo+RNA-seq+2020:2024[pdat]&retmode=json&retmax=15&sort=pub_date
    
  2. Get summaries for found PMIDs:

    code
    WebFetch: https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi?db=pubmed&id=12345,67890,11111&retmode=json
    
  3. Present results with:

Output Format

Present search results as a formatted table or list:

code
## Literature Search Results

**Query:** vitiligo transcriptome

| # | Title | Authors | Journal | Year | PMID |
|---|-------|---------|---------|------|------|
| 1 | Example title... | Smith J et al. | J Invest Dermatol | 2023 | [12345](https://pubmed.ncbi.nlm.nih.gov/12345/) |

Tips

  1. Refine searches - Start broad, then add filters if too many results
  2. Use MeSH terms - More precise than keyword searching
  3. Date filters - Use [pdat] for recent literature reviews
  4. Combine queries - Build complex queries with AND/OR/NOT
  5. Check result count - ESearch returns total count in esearchresult.count

Rate Limits

NCBI allows 3 requests/second without an API key. For higher throughput, users should register for an API key at: https://www.ncbi.nlm.nih.gov/account/settings/

Add API key to requests: &api_key=YOUR_KEY