pubmed-paper-exploration

Paper Deep Exploration

Description

This workflow is used for deep digging after you already have a key paper. You can expand in three directions: similar literature, who this paper cites, and who cites it, then further build a citation tree or research timeline.

Trigger Conditions

• "Related research for this paper"
• "Who cites this paper?"
• "Whose research is this paper built on?"
• "Help me draw the citation tree for this paper"
• A PMID or a clear seed paper title has already been provided

---

If there is no PMID yet

First use unified_search to find the seed paper:

unified_search(
    query="remimazolam randomized controlled trial",
    limit=5,
    ranking="quality"
)

After getting the PMID, proceed to the workflow below.

---

Core Exploration Directions

Tool	Function	Direction
fetch_article_details	Get complete article information	Current paper
find_related_articles	Find articles with similar topics	Horizontal
get_article_references	Find who this paper cites	To the past
find_citing_articles	Find who cites this paper	To the future
get_citation_metrics	View citation influence	Assess importance
build_citation_tree	Build citation network	Visualization
build_research_timeline	Build timeline	Evolution context

---

Most Commonly Used Examples

1. First get article details

fetch_article_details(pmids="30217674")

2. Find similar articles

find_related_articles(pmid="30217674", limit=10)

3. Find who this paper cites

get_article_references(pmid="30217674", limit=30)

4. Find who cited this paper

find_citing_articles(pmid="30217674", limit=20)

5. See which papers are most influential

get_citation_metrics(
    pmids="30217674,35678901,34567890",
    sort_by="relative_citation_ratio"
)

6. Directly build a citation tree

build_citation_tree(
    pmid="30217674",
    depth=2,
    direction="both",
    output_format="mermaid"
)

---

Suggested Workflow

Scenario: expand outward from a key RCT

# Step 1: 當前論文
fetch_article_details(pmids="30217674")

# Step 2: 三個方向同時展開
find_related_articles(pmid="30217674", limit=10)
get_article_references(pmid="30217674", limit=20)
find_citing_articles(pmid="30217674", limit=20)

# Step 3: 對重要 PMIDs 做引用影響力排序
get_citation_metrics(
    pmids="35678901,34567890,33456789",
    sort_by="citation_count"
)

# Step 4: 視覺化引用網路
build_citation_tree(
    pmid="30217674",
    depth=2,
    direction="both",
    output_format="cytoscape"
)

---

If you want to see temporal evolution

Build a timeline with a single PMID or a set of PMIDs

build_research_timeline(
    pmids="30217674,35678901,34567890",
    topic="Remimazolam Clinical Development",
    output_format="mermaid"
)

Build a timeline by topic

build_research_timeline(
    topic="remimazolam",
    max_events=20,
    output_format="text"
)

---

When you want to read the content in depth

Full text

get_fulltext(pmid="30217674", extended_sources=True)

Figures and tables

get_article_figures(pmid="30217674")

---

Practical Judgment

Find core papers in the field

Prioritize:

• find_related_articles articles that commonly appear in both find_citing_articles and
• get_citation_metrics articles with high RCR or citation count in
• build_citation_tree nodes at the center of the network in

Find the latest developments

Prioritize:

• find_citing_articles recent-year articles in
• build_research_timeline the most recent milestones in

Find foundational literature

Prioritize:

• get_article_references early high-impact papers returned by

---

Common Mistakes

• fetch_article_details is not equivalent to a citation network tool; citations and cited-by should respectively use get_article_references and find_citing_articles
• build_citation_tree accepts only one PMID at a time; do not pass multiple papers at once
• If there is no seed PMID, first use unified_search to find it; do not guess directly