Blat Clumptohsprange Error

Question

[arnauvich]

Hi,

It's my first time on Biostars. I would try to explain my problem the best I can. I'm using BLAT for matching predicted proteins from clustered metagenomics reads (26M of CDSs) against a non-redundant protein database (md5nr). The database have 8 Gb aprox. After running 10 minuts BLAT reports the following error:

 blat: gfBlatLib.c:794: clumpToHspRange: Assertion `range->tEnd <= tSeq->size' failed.
Loaded 7808493714 letters in 21461407 sequences

I cut the database (first 1000 sequences) and it works perfectly. My hypothesis is that there some problem in the database, but I would like to know if someone have ever reported this error and how to fixed. Here is the BLAT code for the error ( http://www.pih.bio.br/references/blatRef/gfBlatLib_8c-source.html#l00718 ).

Well thanks in advance,

Cheers.

Answer 1

[arnauvich]

Ok, I post the answer of my question in case that someone need it in the future.

The error is due the database size. When it is bigger than 4GB, BLAT crash.

Answer 2

[opplatek]

Thanks for this information. I had similar problem but in my case I just received core dumped without any further explanation. Cutting the database into 4GB pieces did the trick.