Although simultaneous analysis of genome-wide SNPs has been popular for over a decade, the problems posed by more SNPs than study participants (more parameters than data points), and correlations among the SNPs (linkage disequilibrium, LD), have not been adequately overcome so that almost all published genome-wide analyses are suboptimal in ways that are easily improved. While there has been much attention paid to the shape of prior distributions for SNP effect sizes, we argue that this attention is misplaced. We focus on what we call the “heritability model”: a low-dimensional, deterministic model for the expected heritability at each SNP, for both individual-data and summary-statistic analyses. The 1-df uniform heritability model has been implicitly adopted in a wide range of analyses, but even within the class of 1-df models there are much better choices. Further, there are many other predictors of heritability based on allele frequency, LD and functional annotations leading to substantial improvements in estimates of heritability and selection parameters over traits, and over genome regions, as well as improvements in gene-based association testing and prediction.
Key collaborators: Doug Speed, Aarhus, Denmark and Melbourne PhD student Anubhav Kaphle.
Professor David Balding
Honorary Professor of Statistical Genetics, the University of Melbourne
David Balding received a BMath from the University of Newcastle (Australia) and a PhD in Applied Probability from the University of Oxford, UK. He then held academic posts in and around London, initially in Mathematics but transitioning through Applied Statistics and Epidemiology/Public Health to Genetics. His research develops mathematical/ computational/ statistical innovations in population, evolutionary, medical and forensic genetics. He has given expert evidence in many court cases, mostly about interpretation of DNA profile evidence which is the topic of his monograph “Weight-of-Evidence for Forensic DNA Profiles” (Wiley, 2nd ed 2015). He has developed methods for the analysis of genome-wide genetic data to help understand the genetic architecture of diseases and other traits. His applied research has also encompassed genetics of purebred dogs and crop production, evolutionary and demographic inferences in humans and other species. After 30 years in the UK, he returned to Australia in 2014 where he was Professor of Statistical Genetics at the University of Melbourne and Director of the Melbourne Integrative Genomics until retiring in 2021. He is lead editor of the Handbook of Statistical Genomics (4th ed 2019) with co-editors Ida Moltke (Copenhagen) and John Marioni (Cambridge).