Is it chaotic? Chicken Linkage Disequilibrium and quantitative trait loci

Abstract

Linkage disequilibrium (LD) underpins genome-wide association studies (GWAS) to map Quantitative Trait Loci Regions (QTLRs). A common yet misled assumption is that the closest sites to a significant marker are the prime candidates as the causative mutation. However, LD blocks are typically fragmented and intermingled. We studied chicken LD of mapped QTLRs for Mareks Disease in numerous chicken populations. We estimated background LD by random sampling of non-syntenic or syntenic marker pairs from different or same chromosomes. We defined LD blocks as a group of markers located on the same chromosome having significant moderate (0.15 [≥] r2 < 0.7) or high (r2 [≥] 0.7) LD, regardless of distance and the number of mixed markers with no LD. Next, we studied QTLR LD by all SNP markers within the QTLRs. We found very complex, nearly chaotic, LD patterns, with fragmented and interdigitated LD blocks. Exceptional LD was found between QTLRs 4-5. LD blocks and protein networks shared by the two QTLRs suggest possible functional relationships. Thus, potential causative candidates could be found beyond non-LD sites, occasionally at a very large distance from a significant marker, and even in another QTLR. Multiple effects might make LD intrinsically chaotic, thus limiting GWAS informativity and repeatability. Chaotic complexity challenges the practical use of LD across populations, and must be accounted for while interpreting genetic mapping studies.