Prioritize cREs for CRISPRi validation

[adamklie]

Context

Final step: select cREs for CRISPRi validation based on all upstream analyses. First pass: manuscripts/ChromBetaNet/scratch/2025_12_22.

Tasks

• Compile candidate cRE list from all analyses
• Prioritize by: strength of evidence, disease relevance, novelty
• Finalize validation set

[adamklie]

Dependencies

• Blocked by Retrain ChromBPNet models at different resolutions #1, Recall peak-to-gene links using Gaulton lab pipeline #2, Identify gene modules correlated with INS secretion score and/or disease #7, Rebuild base-GRNs for SC-beta cell types #13, Repeat and expand variant effect predictions #15 (needs results from all upstream analyses)
• This is the final convergence point of the project

[adamklie]

Notes from project reorganization (2026-03-23)

cRE identification and validation plan filed in manuscripts/beta_cell_networks/bin/README.md:

Goal

Prepare a list of 100 CREs for experimental validation:

• 98 CREs: Differentially accessible & expressed in 3-cytokine condition
• 2 CREs: Linked to the WARS promoter via ChromBPNet models

Required metadata for Han

• Differential accessibility stats (in beta cells)
• Browser sessions: beta cell untreated/treated, BPNet prediction, critical DNA features
• Links to the specific CREs
• Coordinates of TF binding motifs within each CRE (for guide RNA targeting)

Guide RNA design

• Han will perform manual design targeting single unique genomic locations
• Adam to consult the networks group for high-throughput methods

Disease relevance framing

• 98 DARs should be activated CREs
• Can include 2 WARS variants
• T1D risk variants: which are in accessible cREs? Predicted to affect the cRE?
• Proximity of credible set variants to genes and cREs

[adamklie]

Status Update (2026-04-09)

Final convergence point. Blocked on #1, #2, #7, #13, #15. Deferred.