pirl-unc / hitlist / 25186217241
80%

DEFAULT BRANCH: main
Ran
Jobs 1
Files 27
Run time 1min
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
coverage: 70.692% (+0.2%) from 70.511%
25186217241

push

github

web-flow 
v1.30.16: APM perturbation classifier — binary columns + --apm-only filter (#202) (#202)

Closes user request: 'all APM perturbations curated in binary columns
and an overall union of those and a simple CLI way to export the
studies and/or their entries'.

New module hitlist/apm.py classifies a curator-written perturbation
string against 34 categories spanning the antigen-processing
machinery:

- Class-I APM genes (13): B2M, TAP1, TAP2, TAPBP, ERAP1/2, PDIA3,
  CALR, CANX, IRF2, GANAB, SPPL3, NLRC5
- Proteasome subunits (4): PSMB5/8/9/10 (LMP2/LMP7/MECL1 aliases)
- Inhibitor classes (4): proteasome (bortezomib, MG132, epoxomicin,
  carfilzomib, lactacystin, ONX-0914), TAP (ICP47, US6), ERAP
  (DG013A), cathepsin (leupeptin, E-64, pepstatin)
- Class-II APM (5): CIITA, HLA-DM, HLA-DO, CD74, cathepsin
- Class-II loci umbrella (2): RFX, bare-lymphocyte-syndrome
- Cytokine inducers (5): IFN-gamma/alpha/beta, TNF-alpha, LPS
- TAP-deficient cell line lineage (1): T2 / RMA-S

Schema additions:
- generate_ms_samples_table emits one apm_<gene>_perturbed bool per
  category, plus apm_perturbed (union) and apm_genes_perturbed
  (semicolon-joined list of matching keys).
- generate_observations_table propagates apm_perturbed and
  apm_genes_perturbed through the per-sample join, so peptide-row
  consumers can pivot on perturbation state without re-deriving.

Filters:
- generate_ms_samples_table(apm_only=True) → APM-perturbed samples
- generate_observations_table(apm_only=True) → peptide rows from
  APM-perturbed samples
- hitlist samples --apm-only
- hitlist export observations --apm-only

Real-data sanity (current corpus, before any new curation):
- 246 APM-perturbed samples across 30 studies
- Top genes: ifn_gamma=140, hla_dm=56, erap1=26, tap1=16, canx=14
- Zero matches for the inhibitor categories (curation gap — the
  vocab covers them so future studies light them up automatically)

Defensive parsing:
- Restricts matching to the curator-written ``condition`` a... (continued)

3618 of 5118 relevant lines covered (70.69%)

0.71 hits per line

Coverage Regressions

Lines Coverage ∆ File
102
81.64
 0.17% export.py
91
55.25
 0.42% cli.py
Jobs
ID Job ID Ran Files Coverage
1 25186217241.1 27
70.69
 GitHub Action Run
Source Files on build 25186217241