Evaluating Population Genomic Candidate Genes Underlying Flowering Time in Arabidopsis thaliana Using T-DNA Insertion Lines.

Population genomic scans have emerged as a powerful tool to detect regions of the genome that are potential targets of selection. Despite the success of genomic scans in identifying novel lists of loci potentially underlying adaptation, few studies proceed to validate the function of these candidate genes. In this study, we used transfer-DNA (T-DNA) insertion lines to evaluate the effects of 27 candidate genes on flowering time in North American accessions of Arabidopsis thaliana. We compared the flowering time of T-DNA insertion lines that knock out the function of a candidate gene obtained from population genomic studies to a wild type under long- and short-day conditions. We also did the same for a collection of randomly chosen genes that had not been identified as candidates. We validated the well-known effect of long-day conditions in accelerating flowering time and found that gene disruption caused by insertional mutagenesis tends to delay flowering. Surprisingly, we found that knockouts in random genes were just as likely to produce significant phenotypic effects as knockouts in candidate genes. T-DNA insertions at a handful of candidate genes that had previously been identified as outlier loci showed significant delays in flowering time under both long and short days, suggesting that they are promising candidates for future investigation.

A note on measuring natural selection on principal component scores.

Measuring natural selection through the use of multiple regression has transformed our understanding of selection, although the methods used remain sensitive to the effects of multicollinearity due to highly correlated traits. While measuring selection on principal component (PC) scores is an apparent solution to this challenge, this approach has been heavily criticized due to difficulties in interpretation and relating PC axes back to the original traits. We describe and illustrate how to transform selection gradients for PC scores back into selection gradients for the original traits, addressing issues of multicollinearity and biological interpretation. In addition to reducing multicollinearity, we suggest that this method may have promise for measuring selection on high-dimensional data such as volatiles or gene expression traits. We demonstrate this approach with empirical data and examples from the literature, highlighting how selection estimates for PC scores can be interpreted while reducing the consequences of multicollinearity.