A statistical approach to mutation detection in zebrafish with next-generation sequencing

The zebrafish is an excellent model organism for forward-genetics, with attributes such as small size, rapid development and straightforward imaging enabling mutagenesis screens for a wide variety of phenotypes. For the majority of these screens over the last few decades, the mutations were mapped using bulk segregant analysis (BSA) to establish approximate chromosomal locations, followed by fine mapping using microsatellite markers on hundreds (or thousands) of embryos. This process is very time consuming despite the large clutch sizes of the zebrafish. Next-generation sequencing (NGS) technologies have drastically improved the speed of this process, but there is no consensus on the best method for performing the BSA and fine-mapping analysis on NGS data. Here we describe a simple statistical approach to this problem using difference-in-homozygosity as a single variable with a normal distribution. This approach was used to accurately map and identify the causative mutation in a zebrafish line with a recessive mineralization disorder