Closed bschilder closed 1 year ago
Redo with more thresholds.
Pooled peaks called with SEACR at 0.1 threshold (top 10% strongest peaks).
Includes both internal (Neurogenomics Lab) and external (previously published) bulk CUT&Tag H3K27ac samples (7 samples total). At 80% precision, recall was 64.2%
Only includes internal bulk CUT&Tag H3K27ac samples (5 samples total). At 80% precision, recall was 59.6%.
Call consensus peaks on all our C&T samples (grouped by mark and cell line). Specifically, merge the bam files and recall peaks using SEACR and MACS2.
See here for other strategies on consensus peak calling: https://github.com/neurogenomics/TIPseq/issues/4