BioGist

Get the biological gist of any research paper. Auto-detects genes, variants, and datasets on any webpage.

BioGist is a research sidebar that scans any webpage for biological entities and provides instant context — no uploads, no server, no account. HOW IT WORKS Visit a paper on PubMed, bioRxiv, Nature, or any scientific website. Open the sidebar and click Scan. BioGist detects 18 types of biological entities on the page. Click any entity for details, database links, and related papers from PubMed. WHAT IT DETECTS BioGist recognizes 18 types of biological entities on any webpage, including genes, variants, drugs, species, and clinical trials. Gene detection uses the full HGNC set of 44,959 human gene symbols for accurate matching with false-positive filtering. DETAIL PANELS Click any detected entity to see rich details from public databases — gene summaries, variant frequencies, clinical significance, citation counts, and open access links. Export citations in APA, Vancouver, Harvard, BibTeX, or RIS format. INLINE PUBMED SEARCH Click "Find Related Papers" in any entity detail to see the top 5 PubMed results directly in the sidebar — titles, authors, journal, and year — without leaving the page. DATABASE LINKS Every entity links out to relevant public databases for deeper exploration. The extension connects to established bioinformatics resources when you click an entity for details. RESEARCH TOOLS - Compare two papers side by side to see shared and unique entities - Co-occurrence matrix showing which entity types appear together across tabs - Persistent entity history with search and type filter across sessions - Batch scan up to 20 URLs at once - Personal annotation notes on any entity - Export as plain text, JSON, Markdown, CSV, or BibTeX with scope and type filters ADDITIONAL FEATURES - Adaptive toolbar that collapses into a menu on narrow sidebars - Multi-tab scanning with per-tab result storage - Pin important entities across sessions - Keyboard shortcuts for sidebar toggle and scan - Right-click context menu for quick lookups - Paste button for scanning text from PDFs - Entity type toggles to show or hide any of the 18 types - Light and dark themes - 24-hour local caching of API results PRIVACY All entity detection runs locally in your browser. No page content is uploaded anywhere. The extension only contacts public bioinformatics APIs when you click an entity to view details. No analytics, no tracking, no account required. Report issues: https://github.com/oriclabs/biolang/issues Built by ORIC Labs — https://oriclabs.com