Augustus
Augustus copied to clipboard
etraining always crashes on one large gene
I've been having problems with my installation of BUSCO not working properly. After a great deal of troubleshooting, I have narrowed the problem down to the etraining binary, which is part of Augustus (v3.3.3). The etraining binary always crashes on me with a "Segmentation fault (core dumped)" error message when it gets to a particular gene. If I remove that gene from the input, etraining runs without a problem.
The problematic gene is quite large (77 kbp) and has a lot of exons (attached below):
I'm running etraining with 32 GB for just that single gene, so I don't think it is a shortage of memory. Any ideas?