Post

2 followers Follow
0
Avatar

Restrict Open Reading Frames by Direction

It would be useful if an additional options were incorporated into the ORF finder that:

1) Allowed you to designate that only ORFs in the 5' to 3' direction were mapped (or 3'-5', or both, as long as we can choose)

2) Incorporated splice sites. Viral genomes have splice sites internal to ORFs of genes (they code in multiple frames, and the genes overlap...clever creatures).  To detect these spliced genes, a system similar to ORF could be incorporated with the following changes:

  1. Allow a nucleotide sequence to be searched.  This would be similar to the "ATG" in the ORF finder, but it would need to be longer.  A wildcard feature would be useful, too, such that the user could designate the number of mismatches within their dictated nucleotide sequence.  For instance, search for the splice site sequence, "GATTTGCA" and allow "2" mismatches.  
  2. Note that the first 'exon' of the gene still begins with an ATG, but does not end with a stop codon.  The downstream exon starts with the splice sequence and ends with a stop codon.

Thanks,

Jake

D. Jake VanBelzen

Please sign in to leave a comment.

3 comments

0
Avatar

Hi Jake,

I can submit (1) to our developers as a feature request.  However, it would be useful to understand why you don't want to see any ORF's on the reverse strand.  If you can justify the general utility of this feature then this would be helpful.

Regarding (2), the ORF Finder, this tool is just that, an ORF finder, finding ORF regions defined by start and stop codons.  What you are asking for is really beyond the scope of this tool.  Reliable and useful prediction of exon splicing, or non-standard viral splicing really requires heuristics such as that used GeneMarkS (http://exon.gatech.edu/GeneMark/).  If you know what your splice site motif is, then you could use Annotate and Predict -> Search for motifs to identify and annotates regions on your sequence that are potential splice sites.  

 

 

 

 

 

Moreland Gibbs 0 votes
0
Avatar

Hi Gibbs,

For (1), when users submit a sequence of known orientation, it is useful to only show ORFs in the expected direction.  Since I have a few small ORFs that I am looking for, I end up getting tons of ORFs in the reverse direction that are simply clutter in my application.

2) Thanks for your recommendation.

Jake

D. Jake VanBelzen 0 votes
0
Avatar

Hi Jake,

For (1), something you might find useful is the annotation filter. You could type in "direction:forward" (without quotes) to only show forward directed annotations. Or if you want to see all other annotation types in either direction but only ORFs in the forward direction, use "direction:forward or not type:orf" (without quotes)

Matt Kearse 0 votes