BLAST - plus/minus strand - non coding region information

[niti217]

Yes. You just install Blast, read the manual very carefully, set up your databases, and run it. Then you learn enough bash so that you can parse your results file.

Thanks again so much. I am excited, can't wait to see the results.

[rhinoceros]

Thanks so much for the useful input. Though I have one more question - to blastx against a protein db, is there an automated way to do for 1800 + genes.

thanks again for your time on this.

Yes. You just install Blast, read the manual very carefully, set up your databases, and run it. Then you learn enough bash so that you can parse your results file.

[niti217]

Well, you could first blastx against a protein db (like nr or refseq_protein or whatever) and choose tabular output. From that you'll parse the information about hit regions. Next you find a clever way to select the non-hit regions from your files and blastn those against some ncRNA db (perhaps one you combine from all available ncRNA dbs?).

Thanks so much for the useful input. Though I have one more question - to blastx against a protein db, is there an automated way to do for 1800 + genes.

thanks again for your time on this.

[rhinoceros]

Well, you could first blastx against a protein db (like nr or refseq_protein or whatever) and choose tabular output. From that you'll parse the information about hit regions. Next you find a clever way to select the non-hit regions from your files and blastn those against some ncRNA db (perhaps one you combine from all available ncRNA dbs?).

[niti217]

Hi

First of all thanks for looking into my post. I need to seek information about non protein coding regions in the reverse strand, using either BLAST or any other tools.

For example if I have a gene ADHB14B (rat species), I use its fasta sequence to run BLAST and then go into the alignments output and trace the plus/minus strands to find out how many non-coding regions/transcripts are there. Then map for each of those transcripts (if non-coding) to find out if they are snoRNA, psuedogenes, miRNA, or any other type of non coding rnas.

Is it possible to automate this process, because I have over ~18000 genes to assess for their noncoding regions on their complementary strand. Its physically very difficult to do for each one of them.

Any inputs on this would be highly appreciated. Thanks once again.