determining library size

[seqgirl123]

How does one go about determining what library size is the best to sequence, for example 200bp libraries versus 250bp or 300bp libraries?

[mollusc]

Library size is determined by read length.If you aim for 2x100 reads,you can go for a 300bp insert which could produce two 100bp non overlapping reads subsequently increasing coverage.

[kjm]

Hi!

I had a few beginner questions about size selection when doing library prep.

Some background info:
-General project is a resequencing project
-Have a reference genome
-Using genomic DNA
-Will be using Illumina Hiseq platform
-Will be looking for SNPs when analyzing

Big question: How do you determine what size your adapters/libraries should be (if doing paired ends)?
-If using beads, your size selection can range from say 200-500 bp. From that range, how do you know which size will be best? I'm guessing this depends on the desired read lengths as stated in the above post?

I think since I will be doing SNP detection, I'm concerned how much a couple hundred bp will affect the outcome.

Hopefully I made some sense in my questions.
Thanks!

[SNPsaurus]

kjm, if you are looking for SNPs there are few things worth considering. When we do a nextRAD SNP discovery project, the main questions are 1) how many markers are needed, and 2) what kind of depth of coverage is needed.

If you are doing whole-genome resequencing, you can't really target a desired number of SNPs. But you'll need a reasonably high coverage to sample both homologous chromosomes in a diploid (~20X), and lower coverage if the genome you are sampling is haploid.

As to the fragment length of the library, in your case it doesn't matter too much. If you want to identify structural variation, then a longer insert size helps. You will want it to be >twice the read length so you don't waste sequencing efforts on redundant information. You might as well keep complexity in the library by not excluding much.

[kjm]

Thanks for the response SNPsaurus. Not sure if it will change anything, but I will be using the SNPs for recombination rate. So if I remember the number correctly, the goal is to have over 10,000 markers (this is the number that keeps popping in my head I think this is right).

[SNPsaurus]

Unless you are re-sequencing closely related small genomes (like yeast), whole genome sequencing will give you many more markers than that. Genotyping-by-sequencing methods were invented for that reason--to sequence just a fraction the genome, allowing more samples to be sequenced in a lane. Check out http://www.ncbi.nlm.nih.gov/pubmed/21681211 for a review.