Wheeler Lab — Conferences

A lightweight, self-updating list of upcoming conferences and workshops relevant to the Wheeler Lab: bioinformatics, ML for biology, genomics, and protein structural modeling.

Live site: https://traviswheelerlab.github.io/conferences/

How it works

• data/conferences.json is the single source of truth. Each entry has a machine-readable start/end (ISO dates, used for sorting and pruning), a human-readable dates_display, plus name, url, location, paper_deadline, and poster_deadline.
• index.html / app.js / style.css render the data as a static page. The page splits entries into Upcoming (end date ≥ today) and a collapsed Past section entirely in the browser — no build step.
• scripts/prune.py removes entries whose end date has passed and refreshes the updated field.

Automation

1. Pruning — .github/workflows/prune.yml runs prune.py weekly (Mondays) and commits the result, so the site never shows finished meetings. It can also be run on demand from the repo's Actions tab.
2. Discovery — a scheduled Claude agent runs every two weeks, searches the web for new relevant conferences within the next ~12 months, adds them to data/conferences.json (de-duplicating on name), and commits directly. See DISCOVERY_PROMPT.md for the exact instructions it follows.

Adding a conference the easy way

Paste the conference's URL (one per line) into add-conferences.txt and commit. On its next run the discovery agent visits each URL, extracts the name, dates, location, and deadlines, adds it to data/conferences.json, and removes the line. To process immediately instead of waiting for the schedule, open the routine at https://claude.ai/code/routines and click Run now.

Editing by hand

Add or correct an entry directly in data/conferences.json and commit. Keep start/end as valid YYYY-MM-DD dates so sorting and pruning work.

Local preview

python3 -m http.server 8000
# then open http://localhost:8000/