Skip to content

DO NOT MERGE until BANC v888 connectivity is live: BANC v626 -> v888 in docs - #526

Merged
Clare72 merged 3 commits into
masterfrom
banc-v888-live
Oct 26, 2026
Merged

Clare72 merged 3 commits into
masterfrom
banc-v888-live

Conversation

@Clare72

@Clare72 Clare72 commented Oct 15, 2026 •

Copy link
Copy Markdown

⚠️ Do not merge until BANC v888 neuron-to-neuron connectivity is live on VFB — i.e. connectome_BANC888_n2n.owl (VFB_connectomics_import#33) has replaced the v626 file on the data server and been loaded into PDB. Until then these pages would describe connectivity VFB does not yet serve.

Stacked on #527 (male-CNS v0.9 → v1.0), which edits the adjacent table rows. Merge #527 first; GitHub then retargets this PR to master.

What

Flips the docs from BANC v626 / Bates2025 to v888 / Bates2026:

Page Change
Data/connectivity.md connectome table: BANC v626 → v888
Data/EM/_index.md dataset table: version v888, Codex link data_version=888
Data/stages.md full-CNS dataset link/short form Bates2025 → Bates2026
Concepts/neuron-counts.md BANC figures for v888 (below)

neuron-counts.md

  • 155,916 is a v888 figure. Bates et al. 2026 (Nature, doi:10.1038/s41586-026-10735-w, full text via PubMed Central PMC13518251) gives 150,841 backbone proofread + 5,075 roughly proofread neurons at the v888 snapshot of 17 April 2026, and never mentions v626. The page previously said the analyses were on v626 with no v626 total published; the row is relabelled v888 and the versions note corrected (adds the paper's 62,773 updated / 42,270 new neurons since the preprint, and the post-snapshot drift: banc_888_meta now gives 155,858 proofread + roughly proofread non-glial neurons).
  • DatasetImages: VFB's Bates2026 has 186,392 images (run_query DatasetImages, count_status: exact, 2026-10-15) → ratio 1.20 (v626 was 143,806 → 0.92). The text keeps v626's under-count as the example of the ratio running the other way, since v888 no longer does.

Before merging

  • Re-run DatasetImages for Bates2026 — the BANC image loader may still be adding images, so 186,392 / 1.20 could move.
  • Male-CNS v1.0 is handled in male-CNS: v0.9 -> v1.0, and cite the published Cell paper #527. Not touched, flagged for separate checking: the "preprint reports 142,719 at v821" clause, which BANC's own metadata docs contradict (they call v626 the preprint materialization) — not verified here.

🤖 Generated with Claude Code

Clare72 and others added 2 commits October 15, 2026 12:55
Flip BANC to v888 / Bates2026 in the connectivity table, the EM dataset
table (version and Codex link) and the stages dataset list. Hold until the
v888 neuron-to-neuron connectivity (VFB_connectomics_import#33) is loaded in
PDB.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
The Nature paper's 155,916 (150,841 backbone proofread + 5,075 roughly
proofread) is the v888 snapshot of 17 April 2026, not a v626 figure: the paper
never mentions v626. Relabel the row v888, correct the versions note, and
update the DatasetImages comparison to VFB's v888 dataset (Bates2026: 186,392
images, ratio 1.20), keeping the v626 under-count as the example of the ratio
running the other way.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
@Clare72
Clare72 changed the base branch from master to male-cns-v1-0 October 15, 2026 11:55
Base automatically changed from male-cns-v1-0 to master October 15, 2026 12:02
@Clare72
Clare72 marked this pull request as ready for review October 26, 2026 09:32
@Clare72
Clare72 merged commit 71c0f24 into master Oct 26, 2026
1 check passed
@Clare72
Clare72 deleted the banc-v888-live branch October 26, 2026 09:35
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant