Most diseases and traits are influenced by multiple genetic factors. However, identifying variants underlying phenotypic differences between individuals is hard and laborious even in model organisms. For instance, Genome-Wide Association Studies (GWAS) require hundreds of C. elegans wild isolates to be individually phenotyped, a process prone to measurement noise, particularly for traits influenced by environmental factors. Alternatively, in Quantitative Trait Loci (QTL) mapping, large recombinant inbred line (RIL) panels need to be generated, maintained and phenotyped, which limits the number of parental genotypes which can be interrogated. Thus, despite the previous success of these strategies in identifying causal loci, a long road still lies ahead for understanding the complex relationship between genotype and phenotype. To gain insights into the genetic basis of complex traits, we developed a novel Bulk Segregant Analysis (BSA) strategy in C. elegans as a versatile and complementary approach to GWAS and panels of recombinant inbred lines (RILs). Our method is based on mating obligate outcrossing strains carrying the
fog-2(
q71) allele for multiple generations, allowing the accumulation of recombination events, and leading to increased mapping resolution. The resulting population is composed of thousands of genetically unique individuals, which can be used to map traits that vary between the parental strains. Furthermore, the population can be directly interrogated at multiple generations for variants affecting fitness, allowing us to study the dynamics of selection. To guide the choice of parameters for our experiments, we developed a computational framework in R, bulkPop, for simulating the BSA pipeline. This framework takes into account the known genetic map and realistic patterns of linkage disequilibrium, and allows us to optimize the number of individuals and number of generations. We implemented our method to study a cross between the reference strain N2 and the wild isolate CB4856 from Hawaii, and found highly reproducible fitness peaks that were closely matched by our simulations. As a proof of principle, we mapped the known
zeel-1/peel-1 genetic incompatibility on Chr. I. and discovered previously unknown loci influencing fitness/growth. In addition, BSA can also be used to map variants affecting gene expression. We used BSA followed by fluorescence sorting of worms (COPAS Biosorter) to map a de novo genetic variant in
sti-1 leading to up-regulation of the chaperone
daf-21 (HSP90). We will further discuss current limitations of the approach and future directions to improve the mapping resolution.