search

Dfam-consortium / RepeatModeler

De-Novo Repeat Discovery Tool
Other
189 stars 22 forks source link

Changing check for larger genomes #67

Closed BioFalcon closed 4 years ago

BioFalcon commented 4 years ago

Hi, I noticed that larger genomes make more database files, which the conditional doesn't account for and therefore fails after successfully building the database. So I changed the conditional to check if exit status is 0.

jebrosen commented 4 years ago

Hi, I noticed that larger genomes make more database files

I remember seeing a report of this but we were unable to reproduce it. What conditions are needed to get rmblast to make a multiple-file database? (e.g. rmblast version, genome size, number of sequences, free disk space or RAM, etc.)

BioFalcon commented 4 years ago

I think the deciding factor for the multiple-file database is the genome size (working with one roughly 30Gbp).

jebrosen commented 4 years ago

I should have asked this right away - do you actually have a multi-part database, as in db.0.nsq, db.1.nsq, and so on - or only db.0.nsq?

BioFalcon commented 4 years ago

Yes, it goes all the way to db.1.nsq

jebrosen commented 4 years ago

I have (manually) merged this new check into the development branch. Thanks for submitting this!