When I run larger 10X raw_feature_bc_matrix the process always be interrupted without any warning

[hajihajilala]

Dear developer:
SEVtras is a great algorithm! Thank you for developing it. I want to use it in my research but when I run larger files using SEVtras.sEV_regconizer the process always be interrupted without any warning and the out_path didn't exist any files. Now my device has 16 cores and 32 threads and 128G memory, I don't know is it enough? I run this algorithm using Jupyterlab in Linux and the input file is the output of cellranger. I have check the reason and I found when the algorithm would spent a lot of time running multi_enrich and the process would suddenly stop work. Are there any time limitation or device limitation?
Thank you very much!

[RuiqiaoHe]

SEVtras was developed based on the size of normal 10X datasets. 128GB is already enough for SEVtras. If you run SEVtras with files much larger than normal 10X, it would take a lot of memory for SEVtras to complete. I suggest you look at the memory usage when the process suddenly interprets. If this is the case, you can subsample the data and run SEVtras again.
If the memory is not an issue, I suggest you run SEVtras in native Linux by command line.