pirl-unc / hitlist / 34309790796
84%

DEFAULT BRANCH: main
Ran
Jobs 1
Files 30
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: 83.723% (+0.002%) from 83.721%
34309790796

push

github

web-flow 
Record why every arm-ambiguous study cannot reach an arm (#366) (#447)

* Record why every arm-ambiguous study cannot reach an arm (#366)

584,966 observation rows sit at pmid_ambiguous or group_ambiguous. Without a
recorded reason each one invites the same investigation again, which is what
#366 asks to stop: a study should be resolved, or confirmed unresolvable from
the deposited data and recorded as such.

arm_resolution records the verdict per study, with arm_resolution_note
carrying the measurement behind it. Every ambiguous row now carries one --
zero unexplained.

  axis_mismatch         255,179   curated arms and recorded metadata are on
                                  different axes.  PMID 33858848 curates per
                                  donor; IEDB records per tissue (29 values)
                                  and never records donor.
  curation_gap          145,279   a per-row discriminator is present and the
                                  arms are not yet curated to use it.  The
                                  only verdict marking real work.
  no_row_discriminator  129,982   measured: cell_name, source_tissue,
                                  antigen_processing_comments and
                                  assay_comments each take exactly one
                                  distinct value across the study.
  multi_arm_evidence     54,526   the evidence positively places the peptide
                                  in more than one arm.  Nothing is missing.

So 75% of the ambiguity is settled -- a property of what was deposited, not
of how carefully anyone curated.

The vocabulary has five values rather than the planned three because the data
showed three genuinely different kinds of unresolvable. Calling 32938616's
case unresolvable would have been wrong: its ambiguous rows carry elution
enumerations naming both arms, so the peptide really was in both.

Two guards stop a verdict becoming a way to stop looking. A note wi... (continued)

6990 of 8349 relevant lines covered (83.72%)

0.84 hits per line

Coverage Regressions

Lines Coverage ∆ File
118
56.32
 -0.03% cli.py
115
90.87
 -0.13% export.py
48
94.56
 0.08% curation.py
17
94.48
 -0.26% qc.py
Jobs
ID Job ID Ran Files Coverage
1 34309790796.1 30
83.72
 GitHub Action Run
Source Files on build 34309790796