About & Methodology

What's actually good in biology

henhao is a scholarly discovery, taste-learning, and reproducibility platform built for experimental and computational biologists across all 26 biology disciplines.

1. Complete Biology Corpus & Journal Linking

We index 363,000+ preprints and peer-reviewed articles across all 25 active bioRxiv biology categories (back to launch in November 2013) plus medRxiv Genetic & Genomic Medicine. Every preprint is continuously linked to its peer-reviewed journal publication and citation metrics via OpenAlex and Crossref, with a two-pass co-author and affiliation ORCID disambiguation graph.

2. Semantic Embeddings, Structured Tags & Verified Full-Text Digests

Every paper is embedded into a 256-dimensional semantic vector space for instant similarity mapping and personalized taste recommendations. Over 185,000 papers are enriched with structured experimental modalities, model organisms, biological processes, and analytical approaches (with hallucinated code/data links stripped by a literal substring guard). Flagship papers feature in-depth full-text digests where every atomic claim is checked against the manuscript by an independent verifier pass.

3. Private-by-Default Taste & Public Reproducibility Ledger

Your personal ratings and reading queue are private by default and immediately train your recommendation feed. Public aggregate scores appear only once at least 3 signed-in readers have rated a paper (and rating histograms unlock at 5 raters), weighted by verified ORCID and institutional status. Our public Reproducibility Ledger lets readers log whether code and protocols run on their environment and compare software dependency diffs.

4. Zotero Integration, RSS Feeds & Bounded Agent / MCP Queries

Zotero / Paperpile / Mendeley Browser Connectors: Every paper page exposes Highwire Press (citation_*) meta tags and Schema.org ScholarlyArticle metadata, plus 1-click BibTeX (.bib) and RIS (.ris) export on paper pages, author pages, search results, and collections.
Direct DOI URLs & RSS: Open any paper via /doi/10.1101/... or subscribe to discipline RSS feeds at /rss.xml.
Model Context Protocol (MCP) & Read-Only Query API: Research agents can connect to POST /api/mcp or read /llms.txt for bounded, rate-limited paper and taxonomy queries (bulk scraping is prohibited).