EMBOSS-FUZZNUC

Question

[Ana]

Hello,

I'm using EMBOSS for the first time and I have several doubts. I have got a short sequence of RNA. I need to find out if there is unintentional hybridisation with RNA regions similar to my sequence (Hybridisation-mediated off-targets).

To do this what I have done is:

1. Installing emboss package on Ubuntu running the following command on terminal:

    sudo apt-get update
    sudo apt-get install emboss
   

2. Download human genome for Searching for potential off-target liabilities in mature human RNAs:

    wget ftp://ftp.ncbi.nlm.nih.gov/refseq/H_sapiens/annotation/GRCh38_latest/refseq_identifiers/GRCh38_latest_rna.fna.gz
   
    gzip -d GRCh38_latest_genomic.fna.gz
   

3. Make an index bd with emboss

    dbifasta -dbname GRCh38_latest_genomic.fna -idformat simple -directory Database directory -filenames newname -release 0 -date date -outfile refMrna1.dbifasta
   

4. To identify potential off-target genes, similarity searches were performed using the pattern match search (fuzznuc):

    fuzznuc -complement -pmismatch 5 -pattern query -sequence refMrna1.fa
   

   I know that the results of that command should show 4 alignments with mature human RNAs, 3 with 100% identity and one with 87.5% identity but what I see in the report is not that, it's a bunch of alignments with 4 or 5 mismatches and none with no mismatches.

I am probably doing something wrong.

Thank you!

[GenoMax]

As you discovered (and in some limited testing I did) using the -pmismatch option literally looks only for that exact number of mismatches. You don't get a range from 0-5 when you specify 5. So your best bet is to run multiple searches and then combine the results.

You may also want to try out blat. It seems to work on UCSC site but I can't get it to work with 20 bp sequences on the command line with the RNA file above.