Capture design for snp sequencing

[Nix]

Has anyone though about how to best design capture probes (we're using Agilent's Custom SureSelect) for sequencing candidate SNPs?

If one designs the probe centered over the SNP then the greatest read coverage with be off either side due to the bimodal distribution of read ends.

Do folks design two capture probes say 100 bp on either side of the target SNP or is this more hassle than it's worth?

[Heisman]

It depends on your read length and if you're doing paired end sequencing. I haven't seen any papers that analyze the most efficient way to design the baits. If possible I think it's a good idea to have at least 2 baits cover any given SNP in case one of them doesn't hybridize efficiently.