Gene model checker

Ask questions about annotation of D. erecta, D. mojavensis, and D. grimshawi projects here.
Post Reply
Posts: 67
Joined: Sun Feb 04, 2007 10:23 pm
Location: California Lutheran University, Thousand Oaks, CA

Gene model checker

Post by drevie » Tue Apr 24, 2012 2:06 pm

One group has a gene on the bottom strand that has the 5' end but not the 3' end. When they plug the data into Gene Model checker, they get an error for the last exon-needs to have a multiple of 3 bases. Since they don't have the last exon(s) of the gene, they don't need a multiple of 3 bases for their last exon. They have clicked on "partial" gene and selected "missing 3' end".

One other note, they originally thought they had the whole gene, but there was no stop codon. They had entered the stop codon coordinates. They then checked on "partial". This closes the stop codon box and lists the "missing" boxes. However, Gene model checker still checks for the stop codon even when the 3' missing box is selected. They changed the last exon coordinate from 36 to 1 and the box where you enter the coordinates indicated an error. We had to change it back to "complete" and delete the stop codon coordinates to fix the problem, then go back to "partial". When the end of the exon was set to 1, there was an error saying there was a noncanonical splice site. Resetting it to 36 fixed that problem to give the "multiples of 3 problem".

What do they do to get it to pass the program?

Posts: 185
Joined: Sun Feb 04, 2007 7:41 pm
Location: Washington University in St. Louis

Re: Gene model checker

Post by wleung » Tue Apr 24, 2012 2:36 pm

In cases where an exon begins within the fosmid and extend beyond the end of the fosmid, you would report the coordinates of the last base of the fosmid in your annotation model. As you have indicated, this would fail the check for the donor site when you submit this as a partial gene model. Irrespective of the configuration options, this type of partial gene models would fail at least one of the checks in the Gene Model Checker. Please provide an explanation of the failure in the project report. The failed item allow us to quickly identify these types of incomplete annotations when we merge the final annotations together.

If you want to pass the Gene Model Checker, you can annotate the model up to the last complete exon that is contained within your fosmid. This is the annotation convention used by the GenBank for partial annotations.

In either case, because the region is covered by overlapping fosmids, the complete version of the partial exon will be found in the annotations of the adjacent projects. Hence we will be able to construct the complete gene model when we combine the annotations together.

Post Reply