Optimal DNA pooling for the detection of single nucleotide polymorphisms

We consider the optimal pooling of DNA to detect single nucleotide polymorphisms (SNPs), sites along the genome at which a population shows variation. The focus is on the detection of low frequency variants. Pooling individuals increases the probability that a rare variant appears in the sample. However, as the pool size increases, the mean number of reads from an individual decreases, making it harder to distinguish reads of a rare variant from errors. A hypothesis test for the detection of SNPs is defined. On the basis of this test, we determine the asymptotically optimal pool size given the parameters of the genome sequencer used, the number of lanes available and a specified significance level.