Hello,

I would recommend you, to run the blast again using output format as tab delimited format. It will be easy for you to parse the results. I would also recommend you, to use word size 7, because you are using short sequences. Hope this will help. 

You could run a blastn against mirBase microRNAs with this command:

$blastProg -db $blastDB -query $inFile -word_size 7 -out $outFile -outfmt "10 qseqid sallseqid qstart qend sstart send qseq sseq length pident nident mismatch gaps evalue" -num_threads $nbThread

(http://www.ncbi.nlm.nih.gov/books/NBK279675/)

The output is a csv file with the filds specified in -outfmt. It's simple to parse. You don't need Bioperl or Biopython to do it.

A candidate microRNA is a sequence that respects some cutoffs like: identity percentage, query coverage, subject coverage, gaps, totale mimatches. and length.

You need to calculate the number of total mismatches:  internal mismatches + external mismatches + internal gaps + external gaps.

internal mismatches and gaps are in the blast output but you need to calculate the external ones.

Thanks a lot Mohamed Remita and Suyash Agarwal for your valuable suggestions

Thanks for the suggestions and it helped me to sort the problem. I have set the option -outfmt 6 and got the Blastn output in tabular form with 12 column headers.