The reason for continual GEITP interest in multifactorial traits, such as human complex diseases including type-2 diabetes (T2D), is that many environmental-toxicant-caused diseases––as well as at least some adverse drug reactions––also qualify as multifactorial traits (meaning ‘phenotypes having contribution from hundreds or even thousands of genes, plus epigenetic effects, plus adverse environmental stressors’). The genetic architecture of common traits such as T2D––including the number, frequency, and effect-sizes of inherited variants that contribute to individual risk––has been long debated.
Genome-wide association studies (GWAS) have identified scores of common variants associated with T2D, but––collectively––these studies explain only a fraction of the heritability of this disease. To test the hypothesis that lower-frequency variants explain much of the remainder contribution to T2D, the Genetics of Type 2 Diabetes (GoT2D) and Type 2 Diabetes Genetic Exploration by Next-generation sequencing in multi-Ethnic Samples (T2D-GENES) consortia [see attached paper] performed whole-genome sequencing in 2,657 European individuals with and without T2D, and exome sequencing in 12,940 individuals, from five ancestry groups.
To increase statistical power, authors expanded the sample size via genotyping and imputation in an additional 111,548 subjects. Variants associated with T2D after sequencing were overwhelmingly common and most fell within regions previously identified by GWAS. Comprehensive enumeration of sequence variation is necessary to identify functional alleles that provide important clues to disease pathophysiology, but this large-scale sequencing project did not support the idea that lower-frequency variants have a major role in predisposition to T2D.
These data seem like a “geneticist’s nightmare” [see attached editorial], because they “fly in the face” of a lot of developing dogma in this field, i.e. contrary to expectation, low-frequency and rare genetic variants do NOT contribute significantly to type-2 diabetes risk(!!). Now, whether this finding will apply to other (or all) multifactorial traits … remains to be seen.
Nature 4 Aug 2o16; 536: 41–47 & News-n-Views pp 37–38
From: Leikauf, George D [mailto:gleikauf@pitt.edu]
Sent: Friday, August 12, 2016 10:06 AM
Subject: RE: Contrary to expectation, low-frequency and rare genetic variants do NOT contribute significantly to type-2 diabetes risk ??
What if one pursues a rare variant having a minor allele frequency (MAF) of 0.5% (i.e. five in 1,000) uncovered in a cohort of ~1300 cases? Would these strong conclusions of this study indicate that this rare variant is not worth studying? ——I would reiterate: this is one (very large cohort) study, and the multifactorial trait being examined is type-2 diabetes (T2D). It’s too early to make any generalizations from this one study of one complex disease. However, this report represents a “heads-up” that large GWAS studies still cannot “find” all of the “missing heritability”, and genomicists are still searching for a complete understanding.
(Note: Extending the analysis to exome sequencing does not explore the regulatory sequence, yet they conclude they have found everything that is contributory.) ——Authors stated that they compared whole-genome sequencing (WGS) in one group with whole-exome sequencing (WES) in another group, as a means to delineate regulatory AND gene-expression variants from gene-expression variants only.
(I would think that environmental influences are actualized through regulatory variants.) ——Nothing is known for certain in this area of research. How about considering the amount of EPIGENETIC effects exerted through environmental stressors?
What is the combined heredity explained by the currently identified polymorphisms? ——Authors repeatedly mention ~45% heritability explained––for a cohort of this size, and the WGS plus WES methodologies that were carried out, and for this complex disease.
Must we depend on –log P as a measure of success? ——I presume you’re referring to the generally accepted P <5.0e–08 (P <5.0 x 10–8) that is used when searching for single-nucleotide variants inside the entire haploid genome that are statistically significantly associated with a phenotype?
Besides P-values, sample size is of course important. In the past we’ve chatted about this. One study that comes to mind {see attached) is this 2o14 PLoS Genet paper on metadata analysis in a cohort of almost 5,000 mice. This is an excellent teaching paper for demonstrating that, by increasing the N by way of combining many studies, one increases the statistical power to detect “significant” gene X environment (GxE) interactions in genome-wide association (GWA) studies. DwN
George