Blat: how to do a good analysis?

[leo_alves]

Hi all,
I'm trying to do an BLASTP-like analysis using BLAT, but I'm not sure if I'm proceeding correct.
I'm working on 64bit-GNU which has Blat installed previously and I didn't prepared an index previously (as my advisor said me).
I want to align my RNA-seq library (~8 million seqs) against an E.coli reference proteome (4583 seqs).
So, I used the follow command:
blat reference.fasta library.fasta output_file
This way produced an empty .psl file (actually, it only has the header).

Alternativelly, by using this command below I had an error message "t and q must both be either protein or dna".

blat -t=prot -q=dna reference.fasta library.fasta output_file
The error appeared when I tried -t=dnax too.

Could anyone help me about how I have to proceed to obtain a correct result?

Very thanks
Leo

[dpryan]

I suspect this email thread answers your question.

[leo_alves]

Hi dpryan,
Thanks for replying
I did the switch they suggested, but it didn't work. I'll look for a tool to translate the reads to protein, so try the other suggestion from theirs.

Best regards

------------
UPDATING

I've just did the translation and the Blating worked!