Merging mixcr scRNA-seq output with Seurat data

[meghatron21]

Hi. I have 10X genomic 5' scRNA-seq fastq files. I ran mixcr like so: ~/mixcr analyze 10x-5gex-full-length --species hsa $sample_R1_001.fastq.gz $sample_R2_001.fastq.gz $output/$sample

I exported clones like so: ~/Programs/mixcr exportClones --split-by-tags Cell -tag cell -cellGroup -uniqueTagCount Molecule $sample.contigs.clns $sample.scRNA.clones.tsv

I see that everything ran without errors, but I'm uncertain on how to map this back to my Seurat dataframe. I tried going through the docs but I'm having trouble. Is there a good resource for this? I essentially want to map each of my cells to likely clonotype.

Thanks, Meghana

[mehdiborji]

@meghanasp21 For now, you have to go through the clones.tsv file, group the file by barcodes, then within each barcode write some clean up function where a dominant chain is picked as the most likely present chain in the cell. In that sense this is more complicated than cellranger's output, but assigning confident chains to cells is not super easy even with cellranger's newest version where you can input both RNA and VDJ data.