How to bridge genome gaps resulting from repeats?

[michaelbarton]

We sequenced a Pseudomonas genome which resulted in several 454 scaffolds. From examining the sequence data it looks like there are many gaps in the sequence because of repeated rRNA genes. Newbler appears to have left these regions as gaps in the sequence because the repeats cannot be resolved.

We are considering using Illumina sequencing to close these gaps, however we are not sure if this will solve the problem either. If we use mate-pair Illumina will this help to bridge these repeated regions? Otherwise what other options do we have as there are many of these repeats and manual PCR is turning out to be laborious and expensive.

[BaCh]

...
We are considering using Illumina sequencing to close these gaps, however we are not sure if this will solve the problem either. If we use mate-pair Illumina will this help to bridge these repeated regions? Otherwise what other options do we have as there are many of these repeats and manual PCR is turning out to be laborious and expensive.

Imagine you have a puzzle (1000 pieces) showing the image of a couple of trees and bushes, but with some patches of blue sky inbetween the trees. No cloud to be seen.

You reconstructed the puzzle except the blue sky parts. Now you want to "close" the missing blue sky areas by using pieces from a puzzle with the same image, but having 10 times more pieces: 10000 pieces.

Ask yourself: how likely is it you will succeed?

Yes, I know: the comparison is slightly flawed. But still.

If you really don't want to do primer walking (difficult enough) and want to close the genome completely, then I'd recommend you add in 454 paired-end sequencing, using a library size of 6kb (might be not enough) to 10kb (should be enough). 20kb if your provider feels that they can deliver a good library construction.

B.

PS: ah, and throw in that Solexa sequencing anyway. It's cheap and will allow you to correct for all those pesky 454 sequencing errors causing fraeshifts in your ORFs.

[michaelbarton]

Thank you for answer, it was very helpful in understanding our problem. Your jigsaw puzzle analogy makes sense. We consider the option of using 454 with a large insert size.