search

JiekaiLab / scTE

MIT License
87 stars 27 forks source link

imposible to read output file #47

Open JavieraVillar opened 1 year ago

JavieraVillar commented 1 year ago

Hello, Thanks for scTE! I did run it in my mouse scRNA seq data, however, I haven't been able to read the output file... it is too big! I have a .csv of 60gb for each of my samples and I can't read them neither using a cluster with 500gb of RAM. As I just want to analyze the expression of transposable elements (specifically the ERVs), I wondered if there is a way to obtain the counts for the TEs only and/or to filter out the genes and TEs with 0 counts, in order to get a smaller output file?

(when I tried to run it with --hd5f True, it didn't work because it went out if memory even with --mem=200G)

thanks a lot for your help! Javiera.

jphe commented 1 year ago

How many threads do you use, scTE takes ~10Gb memory each thread for human and mouse, please maker sure your cluster has enough RAM or set a smaller value for -p option.

If not that reason, it may the format issue. Can you show me you code and a few lines of the input bam file.

JavieraVillar commented 1 year ago

Hello! I dont have any problem running the code. It succeed running. My problem is to read the output file (it is a .csv file because I set --hd5f False) This csv is so big that I can not open it R. I wanted to ask you any suggestions to be able to read this file and add the TEs counts to my seurat object.

thanks again in advance!