isses about gaps in the probes from NimbleDesign

[seraphin]

We aim to use Nimblegen seqcap EZ choice library to target-sequence ~ 100 genes of interests including promoters, coding and intronic regions. NimbleDesign tool provides the information of the probes used for target enrichment. However, when I visualize the probe coverage under UCSC genome browser, several genes contain exons that will NOT be covered by the probes (i.e. gaps).

If I understand the principles of target enrichment correctly, as the sole purpose of the probes is to "enrich" the chromosome regions of interest and all chromosomes are first shattered to be ~500 bp to 1kb in length. Is it correct to assume that if the "uncovered" exons are only a few hundred bases away from the probes, the exons will still be enriched and we'll get the sequence of them.

Any insight to share will be greatly appreciated. Thank you

[ECO]

Depends on your library insert size...and the GC content of the region...most protocols recommend shorter inserts than what you've said.

The coverage profile around a given probe "island" will depend most highly on the insert size. If it's a critical target, I would figure out WHY they can't design probes (repeat elements, GC content, etc)...and perhaps there is some custom design that you could do to get the region.

[seraphin]

Any references/documents explaining why most protocols "recommend" shorter inserts, or is it just because most sequencing machines these days prefer to sequence 50-150 base DNA pieces? It's a surprise to us too that the "exons" would not be covered, as I thought Nimblegen exome seq kits have been quite popular.
ps. I actually tried to contact their tech support, but recently Roche is restructuring their Diagnostics division (esp Sequencing group) and somehow I could not get a hold of their tech support. Anyone else with a similar experience these days?

[Zaag]

Usually the relatively small insert size is because of the clonal amplification step before sequencing.

You do see that probes in the neighborhood of you target help getting the target covered. However this can be difficult with short read platforms.

About roche/nimblegen tech "support"... Well let's say if you have a small target (< 1 Mb) it might be faster to do sangersequencing than to wait for them to answer any question in a meaningful way. If you have a bigger target you might want to consider whole genome More serious: horrible company with good products.