[EMBOSS] showseq and overlapping ORFs

Derek Gatherer d.gatherer at vir.gla.ac.uk
Wed Feb 22 12:03:11 UTC 2006


Hi EMBOSSers

I'm using showseq as follows:

[gath01d at gamma ebv]$ showseq hhv8.fa -format 4 -trans 
105-944,1112-2764,3179-6577,6594-8681,8665-11202,11329-14367,14485-15741,15756-16979,28774-29154,30242-30769 
-out test.showseq -auto

    EMBOSS An error in showseq.c at line 198:
Translation ranges are not in ascending, non-overlapping order.

As can be seen, the failure originates with the subsequence 6594-8681 
slightly overlapping with the next one 8665-11202.

Is there a way round this on the command line or would it require a 
code tweak?  It would be good if there was, since often in viral 
genomes (this is HHV8, as it happens) ORFs are not cleanly "in 
ascending, non-overlapping order" as the program would seem to require.

A related question: all the above are top-strand ORFs, but further 
down there are a few complementary strand ones.  What combination of 
parameters would I use to indicate that I want some translated on the 
top and some on the bottom?  I could of course use format -6 and get 
all six frames, but that is a bit messy for the output I want.  I was 
thinking that maybe it would need to be something like:

showseq hhv8.fa -things B,N,T,S,B,1,A,F -trans 
105-944,1112-2764,3179-6577,6594-8681,8665-11202,11329-14367  -things 
B,N,T,S,B,-1,A,F -trans 
14485-15741,15756-16979,28774-29154,30242-30769 -out test.showseq -auto

ie, using things to specify that for some ORFs I want the translation 
on -1 instead of 1, but the above command just outputs DNA sequence 
with no translation.

Any ideas gratefully appreciated

Derek




More information about the EMBOSS mailing list