Skip to content

My platform is not in the supported list. Can I still use CartLoader?

Yes. Any platform that can produce a transcript-level table works.

1. Prepare a transcript table

A transcript-level CSV/TSV with these fields:

  • X coordinate
  • Y coordinate
  • feature name (typically gene)
  • count (or one row per transcript)
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
cartloader run_together \
  --platform generic \
  --raw-transcript /path/to/input.tsv.gz \
  --colname-transcript-x x \
  --colname-transcript-y y \
  --colname-transcript-feature gene \
  --colname-transcript-count count \
  --units-per-um 1 \
  --out-dir /path/to/out/my_sample \
  --width 18 --n-factor 12

The Generic tutorial walks through this on real data, including a comma-separated file and a cell-ID column. --raw-transcript sends the table through SGE conversion; the --colname-transcript-* options name its columns, and --units-per-um sets how many coordinate units make one micrometer. See the Generic platform page for every option, including a cell-ID column for cell analyses.

Alternative: convert with sge_convert yourself

To control the conversion directly, run sge_convert in generic mode, then continue with the step-by-step commands (run_ficture2, run_cartload2, uploads):

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
cartloader sge_convert \
  --platform generic \
  --in-csv /path/to/input.tsv \
  --csv-delim $'\t' \
  --csv-colname-x x \
  --csv-colname-y y \
  --csv-colname-feature-name gene \
  --csv-colname-count count \
  --units-per-um 1 \
  --out-dir /path/to/out/sge

Notes

  • Start with a small subset first to validate formatting and scaling.
  • If your file has comments, delimiters, or headers that differ from defaults, set the corresponding options explicitly.

See also: