Deconvolution of 3' 10x genomics GEX data with TRUST4

liulab-dfci / TRUST4

TCR and BCR assembly from RNA-seq data

MIT License

286 stars 50 forks source link

Deconvolution of 3' 10x genomics GEX data with TRUST4 #261

Open jasoncrucify opened 7 months ago

jasoncrucify commented 7 months ago

Dear Dev,

I am new user of TRUST4 on single cell TCR seq analysis, regardless, TCR seq could be successfully obtained using bulk RNAseq data.

I wonder, with my 3' 10x genomics GEX data (I have bcl2, fastq (I1, I2, R1 and R2 of 8 different lanes/samples)), bam files), I am wondering if I could deconvolute TCRseq of each single cell from them?

Many thanks for your response! BW,

mourisl commented 7 months ago

It's hard to get TCR from 3' data. This is because the VDJ region is on the 5'-end of the transcript. You can still get some TCRs, but the sensitivity probably will be around 5%, which is difficult to draw any informative conclusions.