Gaps in Solexa Libraries

[inhumataq]

Hi all,

I have made and assembled multiple Illumina libraries before (single and PE), however, the last couple of genomes had some very weird things going on that I have no explanation for. These are bacterial genomes and are approximately 6Mb. When I got the data back and looked at the reads only about half the genome was represented, and there seemed to be islands of high coverage (20x or higher) interspersed with regions with absolutely no coverage (like, 0). Again, this was pretty much random throughout the whole genome so that after final assembly I get thousands of contigs that are hundred(s) of bp long, but with gaps in between them of hundred(s) of bp. I sheared them with a sonicator, but this is not different than what I was using before. This is the first time I used homemade adaptors and primers though, but I got substantial amplification after PCR.

Any thoughts?

[dvh]

There are some proprietary modifications to Illumina's oligos.

[Mark]

Hi all,

I have made and assembled multiple Illumina libraries before (single and PE), however, the last couple of genomes had some very weird things going on that I have no explanation for. These are bacterial genomes and are approximately 6Mb. When I got the data back and looked at the reads only about half the genome was represented, and there seemed to be islands of high coverage (20x or higher) interspersed with regions with absolutely no coverage (like, 0). Again, this was pretty much random throughout the whole genome so that after final assembly I get thousands of contigs that are hundred(s) of bp long, but with gaps in between them of hundred(s) of bp. I sheared them with a sonicator, but this is not different than what I was using before. This is the first time I used homemade adaptors and primers though, but I got substantial amplification after PCR.

Any thoughts?

Some clarification please. Were all of these "last couple of genomes" amped with your homemade primers? Though I have no experience with homemade primers I know others who have used them successfully. Are you assembling your reads with a reference or de novo and with what software? Are you noticing an increase in the number of unassembled reads?

[inhumataq]

The problematic ones were all amped with homemade primers, but amplification is as good as I've gotten with with the kit primers. I have tried a couple of things to ferret out the problem, my first clue as to a problem came when assembly with Edena gave a max contig length of 800bp with 12,000,000 reads. That was a pretty big red flag. So then I went and aligned reads onto sequences from this genome that I have sequenced before with traditional Sanger. It was here that I saw the huge peaks and valleys in coverage (20x to 0x over and over again).

There seems to be an increase in unassembled reads, but the number of bad reads (more than 3 N's) is normal relative to the one's I've done before.