brentlab.org — recovered archive, handoff bundle

This bundle is the static recovered web presence of the Brent group, ready for publication on GitHub Pages at brentlab.org.

What’s in this bundle

The directory you’re reading this from is the publishable site root. Two archives live underneath it:

  • The Brent Lab at Fred Hutchinson Cancer Research Center — original at brentlab.fredhutch.org, lab pages from roughly 2009 to 2024. Lobby: brent/index.html.
  • The Molecular Sciences Institute — original at molsci.org, from roughly 1996 to 2009. Lobby: external/molsci.org/index.html.

The top-level entry page (index.html) frames both archives with cards linking into each lobby. There is also a small WordPress and Squarespace blog archive under external/brentlab.org/systems/ and external/brentlab.net/blog/, surfaced from the Fred Hutch lobby.

Approximately 1,013 HTML pages, ~400 MB total including assets.

Critical: this site MUST be served from the domain root

The archive uses absolute root-relative paths throughout — links like /etc.clientlibs/foundation/clientlibs/main.min.css and /external/molsci.org/index.html. These only resolve when the contents of this directory sit at the root of the served URL space.

  • ✅ https://brentlab.org/index.html
  • ✅ https://brentlab.org/brent/index.html
  • ❌ https://brentlab.org/site/index.html (would break every stylesheet and cross-page link)

In practice that means: put the contents of this directory at the root of the publishing repo (or in /docs configured as the GitHub Pages source). Do not commit this directory as a subfolder.

Publishing files already in place

  • .nojekyll (empty file) — tells GitHub Pages not to run Jekyll on the archive. Without this, paths starting with _ would be hidden.
  • CNAME — contains the single line brentlab.org. Tells GitHub Pages this is the custom domain.

Both belong at the repo root next to index.html.

DNS setup (once)

GitHub Pages knows to serve brentlab.org because of the CNAME file, but DNS for brentlab.org itself has to point at GitHub. At the apex domain, add four A records:

185.199.108.153
185.199.109.153
185.199.110.153
185.199.111.153

Optionally, a www.brentlab.org CNAME record pointing at <github-username>.github.io.

GitHub’s docs: https://docs.github.com/en/pages/configuring-a-custom-domain-for-your-github-pages-site

DNS propagation usually takes minutes to hours. After DNS verifies, GitHub Pages will automatically provision a Let’s Encrypt TLS certificate; that adds another few minutes.

Things in the archive that look like bugs but are intentional

The H5N1 missing-documents notice

Across four pages in brent/en/center-biological-futures/ — document-repository.html, documentrepository.html, viral-library.html, and virallibrary.html — 196 links that originally pointed at policy PDFs about the 2011 H5N1 transmissibility research now point at brent/en/center-biological-futures/notice-missing-documents.html. The notice explains that those specific documents were not preserved in this archive. Visible link text on each page was left unchanged. This is deliberate, not a recovery failure.

18 references to a missing ResearchResourcesPage.html

The original Fred Hutch site had a page at /brent/en/ResearchResourcesPage.html that was referenced from 18 other pages but was not recovered (it appears to no longer exist on the live host either). Those 18 references remain in place as broken links pointing at the missing path, on the chance that a future recovery effort fills it in. Do not “fix” them by rewriting or deleting them.

MSI redirect stubs

external/molsci.org/www.molsci.org/Redirect*.html (15 small files) are the original CMS’s “this URL has moved” placeholders. Their meta-refresh URLs were rewritten to point at the MSI archive lobby instead of the now-dead www.molsci.org/ root. So they will redirect visitors arriving via old bookmarks into the archive rather than out to a 404.

First-publish spot-check list

After deploying, please verify the following pages render correctly with their original CSS and that internal links work:

  1. https://brentlab.org/ — top-level index, both cards render with styling
  2. https://brentlab.org/brent/index.html — Fred Hutch lobby, full nav
  3. https://brentlab.org/external/molsci.org/index.html — MSI lobby
  4. https://brentlab.org/brent/en/research.html — original FH page, verifies the absolute-path CSS chain
  5. https://brentlab.org/brent/en/center-biological-futures/document-repository.html then click any link in the document list — should land on the notice-missing-documents page
  6. https://brentlab.org/external/molsci.org/www.molsci.org/Dispatch/index.html — a sample MSI Dispatch page
  7. https://brentlab.org/external/brentlab.org/systems/index.html — the WordPress “Systems” blog era

If any of those load with broken styling, almost certainly the absolute path issue described above — verify the bundle’s contents are at the served root.

Origin of this recovery

The archives were reassembled in May 2026 by Claude (Opus 4.7) working from web-archive snapshots and other publicly available copies. The reassembly is read-only: nothing here was authored, only recovered. Where cross-references could not be resolved they were left as broken pointers rather than rewritten or invented.

Questions about the recovery itself: ask Roger.