Chrome Extension PubMed Search Helper: Build Your Own Research Tool

PubMed remains the primary research database for biomedical literature, but its interface can feel dated and cumbersome for power users managing multiple searches. Building a Chrome extension to enhance PubMed search functionality gives you complete control over your research workflow. This guide walks through creating a practical PubMed search helper extension from scratch.

Why Build a PubMed Search Helper

The default PubMed interface requires repeated navigation through filters, manual entry of search terms, and separate steps to access saved searches. A custom Chrome extension can consolidate these operations into a single popup or side panel. Researchers who conduct literature reviews across multiple sessions particularly benefit from automation that preserves search history and provides one-click access to frequently used queries.

The PubMed E-utilities API provides programmatic access to search results, citations, and metadata. This API serves as the backend for any extension that fetches data beyond what the web interface displays.

Extension Architecture Overview

A Chrome extension for PubMed search assistance typically consists of three components:

Manifest file — Defines permissions, browser action, and extension metadata
Popup or side panel — Provides the user interface for search input and results
Background script — Handles API calls and long-running tasks

The manifest version 3 format represents the current standard. Here is a minimal manifest configuration:

{
  "manifest_version": 3,
  "name": "PubMed Search Helper",
  "version": "1.0",
  "description": "Enhanced PubMed search with custom filters and saved queries",
  "permissions": ["storage", "activeTab"],
  "host_permissions": ["https://eutils.ncbi.nlm.nih.gov/*"],
  "action": {
    "default_popup": "popup.html",
    "default_icon": "icon.png"
  }
}

The host_permissions field grants the extension access to the PubMed E-utilities API domain. The storage permission enables persisting saved searches across browser sessions.

Building the Search Interface

The popup HTML provides the primary interaction point. A clean interface includes a search input field, filter options, and a results container:

<!DOCTYPE html>
<html>
<head>
  <style>
    body { width: 320px; padding: 12px; font-family: system-ui; }
    .search-box { width: 100%; padding: 8px; margin-bottom: 8px; }
    .filters { display: flex; gap: 8px; margin-bottom: 12px; }
    .filter-btn { flex: 1; padding: 6px; cursor: pointer; }
    .filter-btn.active { background: #0066cc; color: white; }
    #results { max-height: 300px; overflow-y: auto; }
    .result-item { padding: 8px; border-bottom: 1px solid #eee; }
  </style>
</head>
<body>
  <input type="text" id="searchInput" class="search-box" placeholder="Search PubMed...">
  <div class="filters">
    <button class="filter-btn active" data-filter="all">All</button>
    <button class="filter-btn" data-filter="review">Reviews</button>
    <button class="filter-btn" data-filter="clinical">Clinical</button>
  </div>
  <div id="results"></div>
  <script src="popup.js"></script>
</body>
</html>

The filter buttons apply pre-configured search modifiers. PubMed supports specialized filters like [review] for review articles and [clinicaltrials] for clinical trial publications.

Implementing the Search Logic

The popup JavaScript handles user interactions and communicates with the PubMed API. The E-search endpoint returns PMIDs matching your query:

document.getElementById('searchInput').addEventListener('keypress', async (e) => {
  if (e.key === 'Enter') {
    const query = e.target.value;
    const activeFilter = document.querySelector('.filter-btn.active').dataset.filter;
    const modifiedQuery = applyFilter(query, activeFilter);
    const pmids = await searchPubMed(modifiedQuery);
    displayResults(pmids);
  }
});

function applyFilter(query, filter) {
  if (filter === 'review') return `${query} AND review[pt]`;
  if (filter === 'clinical') return `${query} AND clinical trial[pt]`;
  return query;
}

async function searchPubMed(query) {
  const baseUrl = 'https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi';
  const params = new URLSearchParams({
    db: 'pubmed',
    term: query,
    retmode: 'json',
    retmax: '20',
    sort: 'relevance'
  });
  
  const response = await fetch(`${baseUrl}?${params}`);
  const data = await response.json();
  return data.esearchresult?.idlist || [];
}

The esearch.fcgi endpoint returns PMIDs (PubMed IDs) as an array. Each PMID can then fetch detailed article information through the esummary or efetch endpoints.

Adding Citation Formatting

One of the most useful features for researchers is automatic citation generation. The extension can format citations in multiple styles:

async function fetchArticleDetails(pmid) {
  const url = `https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi?db=pubmed&id=${pmid}&retmode=json`;
  const response = await fetch(url);
  const data = await response.json();
  return data.result[pmid];
}

function formatCitation(article, style = 'apa') {
  const { title, authors, source, pubdate } = article;
  const authorList = authors?.author?.map(a => a.name).join(', ') || 'Unknown';
  
  if (style === 'apa') {
    return `${authorList} (${pubdate}). ${title}. ${source}.`;
  }
  if (style === 'bibtex') {
    const key = authors?.author?.[0]?.name?.split(' ').pop() || 'unknown';
    return `@article{${key}${pubdate?.split(' ')[0]},
  author = {${authorList}},
  title = {${title}},
  journal = {${source}},
  year = {${pubdate?.split(' ')[0]}}
}`;
  }
  return `${authorList}. ${title}. ${source}. ${pubdate}`;
}

Storing the selected citation style in Chrome storage ensures the preference persists:

chrome.storage.sync.set({ citationStyle: 'apa' });
chrome.storage.sync.get(['citationStyle'], (result) => {
  const style = result.citationStyle || 'apa';
  // Apply style to citations
});

Saving Searches for Quick Access

The storage permission enables saving frequent searches. This approach works well for literature reviews that require recurring queries:

function saveSearch(query, label) {
  chrome.storage.sync.get(['savedSearches'], (result) => {
    const searches = result.savedSearches || [];
    searches.push({ query, label, date: new Date().toISOString() });
    chrome.storage.sync.set({ savedSearches: searches });
  });
}

function loadSavedSearches() {
  chrome.storage.sync.get(['savedSearches'], (result) => {
    const container = document.getElementById('savedSearches');
    (result.savedSearches || []).forEach(search => {
      const btn = document.createElement('button');
      btn.textContent = search.label;
      btn.onclick = () => {
        document.getElementById('searchInput').value = search.query;
        document.getElementById('searchInput').dispatchEvent(new KeyboardEvent('keypress', { key: 'Enter' }));
      };
      container.appendChild(btn);
    });
  });
}

Loading Saved Searches on Startup

Initialize the saved searches when the popup opens:

document.addEventListener('DOMContentLoaded', () => {
  loadSavedSearches();
  setupFilterButtons();
});

Extension Deployment

Once development completes, package the extension through Chrome’s developer dashboard or load it locally for testing:

Navigate to chrome://extensions/
Enable Developer mode (toggle in top right)
Click Pack extension and select your extension directory
Note the generated .crx file for distribution

For testing without packaging, use the Load unpacked option to select your extension directory directly.

Extending the Extension

Several enhancements expand the utility further:

Abstract fetching — Retrieve and display article abstracts directly in results
Citation export — Export selected citations to BibTeX or RIS format for reference managers
Search history — Track and visualize search patterns over time
Keyboard shortcuts — Bind global shortcuts for quick popup access

The PubMed E-utilities API supports additional endpoints for fetching full records, checking citation counts, and accessing related articles. These features transform a basic search helper into a comprehensive research tool.

Built by theluckystrike — More at zovo.one*