Use cellranger GEX format as input to secondary analysis

[yuxuanlan]

Description of feature

Cellranger outputs

matrix.mtx, 
barcodes.tsv,
eatures.txv

are currently not accepted by the secondary analysis.

It would be nice to have the above format, or the .gz version, accepted. Thank you!

The workaround I used:
convert from cellranger format to h5ad format

ingularity exec /ei/software/testing/eisca/sifs/scanpy_1.7.2--pyhdfd78af_0.sif \ $root/Software/eisca/1.4_YL_fork/bin/mtx_to_h5ad.py \
  --aligner cellranger --verbose True --txp2gene '' --star_index '' \
  --input_data $mtx_dir/matrix.mtx \
  --barcode $mtx_dir/barcodes.tsv \
  --feature $mtx_dir/features.tsv \
  --sample 10kHumanPBMC \
  --out ./run2.secondary/10kHumanPBMC_matrix.h5ad

and run pipeline

source /ei/software/testing/eisca/eisca-1.0_IPG && \
nextflow run $root/Software/eisca/1.4_YL_fork/main.nf \
    --analyses secondary \
    --h5ad ./run2.secondary/10kHumanPBMC_matrix.h5ad \
    --input Samplesheet.10kHumanPBMC.csv \
    --genome GRCh38 \
    --outdir $analysis_dir/results \
    -profile singularity --aligner cellranger --protocol 10XV3 \
    --args_qccellfilter '--min_genes 50' --args_clustering '--integrate harmony' \
    -with-report -resume

This was tested on commit d5388d1 in Nov 2024.