huangyh09 / brie

BRIE: Bayesian Regression for Isoform Estimate in Single Cells
https://brie.readthedocs.io
Apache License 2.0
41 stars 15 forks source link

some questions about psi distribution #53

Open TuoCai2000 opened 1 year ago

TuoCai2000 commented 1 year ago

Hello!Thank you for developing Brie! I am interested in alternative splicing events, so I applied BRIE2 to my research. I have eight 10x genomics samples which belong to four groups. I used brie-count for eight times and got eight brie_count.h5ad files for each sample , then I merged them together into one brie_count.h5ad file to run brie-quant(mode2-quant) ,I am wondering if I am handling the data the right way? After that , I got a huge h5ad file with 93357 cells and 4266 splicing events. I made four density plots using ggplot2 about their psi distributions but found them seem to have the same distribution ( massive cells psi values close to 0, while other cells close to 1). I am not sure they are right , I am wondering have you ever seen a psi distribution like this before? I am looking forward to your advise. Thank you so much ! Tuo