ESPE Yearbook of Paediatric Endocrinology

TOPMed Anthropometry Working Group; NHLBI Trans-Omics for Precision Medicine (TOPMed) Consortium et al. Institute for Molecular Bioscience, University of Queensland, Brisbane, Queensland, Australia.Nat Genet. 2022 Mar;54(3):263-273.Abstract: https://pubmed-ncbi-nlm-nih-gov.proxy.kib.ki.se/35256806/

In brief: Genome-wide association studies (GWAS) on unrelated individuals rely on the correlation of common SNP polymorphisms with nearby causal gene variants. In contrast to classical pedigree-based studies, SNP-based approaches only capture a part of the heritability of traits and diseases. This study estimated the heritability of height and BMI based on whole-genome sequence data and identified insufficient SNP tagging of rare variants in low linkage disequilibrium as a main cause for the missing heritability in common SNP approaches.

Commentary: With the increasing number of genomes available for research, GWAS has become a key method to identify genetic variants associated with diseases and phenotypes. Nevertheless, the mapping of all common SNPs as used in current studies is still unable to explain the heritability predicted by pedigree based studies. While this decreased sensitivity is well-know, hypotheses on the reasons for this evidence gap have not been proven so far.

Wainschtain et al used a large number of genomes with European ancestry for SNPs and insertion-deletions, and estimated a SNP-based heritability of 0.5-0.56 for adult height. In order to identify the causes for the gap to the expected total estimate of 0-7-0.8, grouping of minor allele frequencies and linkage distance was performed, increasing heritability estimates for height to 0.70. This discrepancy could mainly be explained by rare variants (low MAF) in low linkage distance. Further, the group investigated the enrichment in heritability for high or low impact on the protein and showed that rare, low-LD and protein-altering variants are more enriched for trait heritability than non-coding variants.

This highly sophisticated approach impressively shows a potential methodological cause for the reported missing heritability of traits and diseases in SNP-based GWAS studies. Importantly, the identified rare variants in low linkage distance could only be covered by use of whole genome sequencing (WGS) due to their lack of coverage in SNP arrays. Especially for polygenic diseases, this study strongly supports the use of WGS in large sample sizes for increased potential to detect causal rare variants.