Quantile-based permutation thresholds for QTL hotspot analysis: a tutorial

QTL hotspots, groups of traits co-mapping to the same genomic location, are a common feature of genetical genomics studies. Genomic locations associated with many traits are biologically interesting since they may harbor influential regulators. Nonetheless, non-genetic mechanisms, uncontrolled environmental factors and unmeasured variables are capable of inducing a strong correlation structure among clusters of transcripts, and as a consequence, whenever a transcript shows a spurious linkage, many correlated transcripts will likely map to the same locus, creating a spurious QTL hotspot. Permutation approaches that do not take into account the phenotypic correlation tend to underestimate the size of the hotspots that might appear by change in these situations (Breitling et al. 2008). This issue motivated the development of permutation tests that preserve the correlation structure of the phenotypes in order to determine the significance of QTL hotspots (Breitling et al. 2008, Chaibub Neto et al. 2012). In this tutorial we present software tools implementing the NL-method (Chaibub Neto et al. 2012), the N -method (Breitling et al. 2008), and the Q-method (West et al. 2007, Wu et al. 2008) permutation approaches.