MNR in High quality discrepancy regions

Ask questions about sequence improvement / finishing D. mojavensis projects here.
Post Reply
drevie
Posts: 67
Joined: Sun Feb 04, 2007 10:23 pm
Location: California Lutheran University, Thousand Oaks, CA

MNR in High quality discrepancy regions

Post by drevie » Wed Feb 25, 2015 4:48 pm

We are finishing DELE contigs. We have some regions where there are lots of HQD that also have MNR. Do we fix the consensus in the whole region, or just the repeats (T's in the picture-94620 and beyond).

We have one or more other regions (other contigs) where the consensus and reads don't agree for hundreds of bases-presumably parts of these have MNR. Do we just tag these regions? or?
Attachments
MNR HQD.png
DELE MNR and HQD
MNR HQD.png (163.84 KiB) Viewed 3054 times

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

Re: MNR in High quality discrepancy regions

Post by wleung » Wed Feb 25, 2015 5:03 pm

This type of discrepancies typically appears after you have removed reads (or did a tear and join) from a region. In some cases, Consed did not recalculate the consensus correctly and it uses a much lower quality read in the region as the consensus. These regions would need to be resolved prior to project submission. One way to resolve these regions would be to tear the discrepant region from the assembly and then run a Miniassembly on the new contig so that the consensus is consistent with most of the reads. You can then join the new contig back into the main scaffold.

drevie
Posts: 67
Joined: Sun Feb 04, 2007 10:23 pm
Location: California Lutheran University, Thousand Oaks, CA

Re: MNR in High quality discrepancy regions

Post by drevie » Thu Feb 26, 2015 12:10 am

I didn't think about problems with forced joins-I just assumed it was elsewhere. I'll ask them about that. Thanks.

Post Reply