Evolutionary Triangulation to Refine Genetic Association Studies of Spontaneous Preterm Birth
19 April 2017
20 April 2017
22 May 2017 (eFirst)
Objective The objective of this study was to apply evolutionary triangulation, a novel technique exploiting evolutionary differentiation among three populations with variable disease prevalence, to spontaneous preterm birth (PTB) genetic association studies.
Study Design Single nucleotide polymorphism (SNP) allele frequency data were obtained from HapMap for CEU, GIH/MEX, and YRI/ASW populations. Evolutionary triangulation SNPs, then genes, were selected according to the overlaps of genetic population differences (CEU = outlier). Evolutionary triangulation genes were then compared with three PTB gene lists: (1) top maternal and fetal genes from a large genome-wide association study of PTB, (2) 640 genes from the database for PTB, and (3) 118 genes from a recent systematic review. Empirical p-values were calculated to determine whether evolutionary triangulation enriched for putative PTB associating genes compared with randomly selected sample genes.
Results Evolutionary triangulation identified 5/17 maternal genes and 8/16 fetal genes from PTB gene list 1. From list 2, 79/640 were identified by CEU_GIH_YRI evolutionary triangulation, and 57/640 were identified by CEU_ASW_MEX evolutionary triangulation. Finally, 20/118 genes were identified by evolutionary triangulation from gene list 3. For all analyses, p < 0.001 except CEU_ASW_MEX analysis of list 3 where p = 0.002.
Conclusion Genes identified in prior PTB association studies confirmed by evolutionary triangulation should be prioritized for further genetic prematurity research.
Evolutionary triangulation, a novel bioinformatics approach, provides independent support for multiple genes previously associated with PTB and presents an alternate filtering metric for genetic analyses using evolutionary history.
This study was presented in part at the Society for Maternal Fetal Medicine's 37th Annual Meeting 2017 (Las Vegas, NV), as an oral concurrent presentation (1/26/17), final abstract ID #11.