pirl-unc / hitlist / 25173064216

Build # 25173064216

Build Type

push

github

Committed by web-flow

Commit Message

v1.30.12: qc discrepancies --by sample for per-cell_name rollup (#197)

Adds a per-sample aggregation level to the discrepancies report.
Default ``--by pmid`` keeps the v1.30.9 behaviour (one row per
study/class). New ``--by sample`` groups by
(pmid, mhc_class, cell_name) so a curator can spot per-sample issues
like "the K562-A0201 transfectant has 30% suspect rows but its
K562-B0702 sibling has 0%". Rows without a cell_name fall back to
the placeholder "(no cell_name)".

API: discrepancies(by="pmid"|"sample"). CLI: --by pmid|sample.

Implements the user's "lots of short MHC II ligands or lots of long
MHC I in some sample" use case directly — same suspect-score
sorting, same severity heuristic, same column shape, just one extra
group key.

Coverage Stats

3534 of 5027 relevant lines covered (70.3%)

0.7 hits per line

Source File

Press 'n' to go to next uncovered line, 'b' for previous

Source Not Available