Hacktoberfest 2026: le issue che i maintainer hanno segnato per ottobre, aperte e adatte ai principianti. Sfoglia le issue Hacktoberfest

CellRanger-compatible `outs/` output (`metrics_summary.csv`, `.h5` matrices, `web_summary.html`) — checking the development method before opening PRs

Aperta
  • Qualcosa non va?
#340 1 commento 1 reazione 1 assegnatario Vedi su GitHub

I maintainer di solito rispondono entro 4 giorni

@Psy-Fer ci sta già lavorando.

Dal 28/10/2026.

Valutazione

Difficoltà
5/5
Tempo stimato
Più di una settimana
Idoneità per principianti
20/100
Tipo di issue
Funzionalità
Chiarezza
Specificata chiaramente
Stato di attività
Attiva
Stack tecnologico
rust
Ambito
backend, data

Direzione di ricerca

Start by reading CONTRIBUTING's policy on compatibility work and review the proposed three PR themes: metrics and gzipped matrices, HDF5 matrices, and the HTML report. The issue names no repository files or tests to run. Done means maintainers decide whether the observed-output method is acceptable, whether the output belongs in this project, and whether the proposed split is right.

Scritto dal modello di indicizzazione a partire dal testo della issue.

Descrizione

sensitive license

CONTRIBUTING says a contribution that reproduces the behaviour of a tool under a
sensitive licence needs to be discussed and agreed before work starts. I have
written this code already, which is my mistake — I am raising it now rather than
opening the PRs, and I will drop the work if the answer is no.

What I would like agreed is the method, which I believe is the one the policy
describes as acceptable, and then whether the result is wanted at all.

What it does

An opt-in flag (--soloCellRangerOuts Yes, off by default, no effect on any
existing output) that writes, beside Solo.out/:

outs/metrics_summary.csv
outs/raw_feature_bc_matrix/{barcodes,features}.tsv.gz, matrix.mtx.gz
outs/filtered_feature_bc_matrix/{barcodes,features}.tsv.gz, matrix.mtx.gz
outs/raw_feature_bc_matrix.h5
outs/filtered_feature_bc_matrix.h5
outs/web_summary.html

The point is that an existing pipeline that reads outs/filtered_feature_bc_matrix.h5
keeps working. Solo.out/ is never altered.

How it was developed — this is the part I want checked

Entirely from observed outputs, which CONTRIBUTING lists as an acceptable
method ("from public documentation, published papers, or observed inputs and
outputs"):

  • The field names, their order, and the number formatting in
    metrics_summary.csv were read from two real CellRanger output files
    (one 8.0.1, one 10.0.0).
  • The HDF5 layout was read from real .h5 outputs with h5py — group and
    dataset names, dtypes, the CSC convention, and the root attributes.
  • The -1 GEM-group barcode suffix was read from real barcodes.tsv.gz.

No CellRanger source was read, ported or paraphrased, by me or by any tool.
In particular I deliberately did not open the JavaScript bundle inside a real
web_summary.html; the page in this work is laid out and drawn from scratch and
is not intended to resemble CellRanger's report. The only thing taken from their
side of web_summary.html is the set of metric names, which are the same ones
already in metrics_summary.csv.

What is and is not claimed

Every figure is rustar-aligner's own. The names are CellRanger's so that
existing parsers keep working; the definitions are not claimed to equal
CellRanger's, and this is stated in DIVERGENCE.md and in the generated page
itself. Two concrete examples of the difference, measured on the same 10M reads:

  • Reads Mapped Confidently to Genome: rustar-aligner reports STAR's uniquely
    mapped fraction (75.1%, against STARsolo's 75.3% on the same reads).
    CellRanger reports 80.8% for the same-named field, so it means something else.
  • Mean Reads per Cell: CellRanger divides all reads by the cell count;
    STAR's Summary.csv divides reads in called cells by it
    (meanReadPerCellUnique = nReadInCellsUnique / nCells,
    SoloFeature_cellFiltering.cpp). The CellRanger-named file uses CellRanger's
    definition; Summary.csv keeps STAR's.
Questions
  1. Is the observed-outputs method acceptable here, given it happened before this
    discussion rather than after?
  2. If so, is this output wanted in rustar-aligner at all, or is it better as a
    separate tool?
  3. If wanted, is the split below right? Each is independently useful and I would
    open them in order.
Proposed PRs
Theme
1 --soloCellRangerOuts: metrics_summary.csv + the gzipped matrix mirror
2 the two .h5 matrices (needs the dependency decision in the linked issue)
3 web_summary.html

Two unrelated pieces from the same work are already separable and do not touch
any of this, so I will open them regardless of the answer here: an EmptyDrops
Good-Turing fidelity fix, and barcode-setup validation.

Those two are pushed and independently reviewed, if it helps to see the standard
the CellRanger-compat work would be held to:
iandriver/rustar-aligner@fix/emptydrops-ambient-species and
@feat/solo-setup-validation, both based on 509129a.

Lingua principale
Rust
Stelle
91
Fork
9
Merge medio
16h 43m
PR unite (30g)
22

Preparare l'ambiente

Come iniziare

  1. Leggi tutta la issue e poi la guida ai contributi del progetto.
  2. Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
  3. Fai un fork del repository e lavora su un branch.
  4. Apri una pull request che faccia riferimento al numero della issue.

Altre issue di scverse/rustar-aligner

Tutte le issue di scverse/rustar-aligner

Issue simili

Altre issue su Rust

Ricevi le nuove issue nella tua casella

Un breve riepilogo di issue GitHub adatte ai principianti.