jessieren / DeepVirFinder

Identifying viruses from metagenomic data by deep learning
Other
116 stars 32 forks source link

DeepVirFinder gets stuck on some sequences #27

Open runphaster opened 3 years ago

runphaster commented 3 years ago

I have a couple dozen sequences that were in my input files that cause the code to get stuck. The load average goes to 0, and progress moving thru the input file ceases. I would like to forward my collection of sequences it gets stuck on so the problem can be debugged/identified going forward.

444thLiao commented 2 years ago

I also get the same situation. I guess mostly is a problem of multiprocessing since it usually gets stuck when the second time of 'processing line' was called.
Which mentioned should be located at nearly the line 183.

SDquest commented 2 years ago

It happened in my running too. I checked and found the reason is that DeepVirFinder can't handle sequences longer than 2.1Mb. I tried removing sequences longer than 2.1Mb in my input, and then DeepVirFinder works. I also tried inputting only one sequence with 2.8Mb length, DeepVirFinder got stuck for a few hours and can't predict it. When I cut this same sequence into two smaller subsequences (2.1Mb and 0.7Mb), DeepVirFinder predicted it successfully in a few minutes.