Friday, December 21, 2018

Last 2018 T2DKP release, dedicated to Todd Green

Todd Green
Today's release in the Type 2 Diabetes Knowledge Portal of multiple new datasets, including associations for 17 new phenotypes, is dedicated to the memory of Todd Green, a Portal team member who passed away unexpectedly on November 19. Todd was an integral part of the Portal project since its inception several years ago, and had worked at the Broad Institute for many years previously. His expertise in applying statistical methods to GWAS and sequencing studies earned him co-authorship on more than 50 papers on the genetics of complex traits, focusing on inflammatory bowel disease and type 2 diabetes. The Portal team will miss him greatly, and we will carry on his spirit in the work that we do.

This release adds to the T2DKP nine new datasets:
  • ADIPOGen GWAS, a study by Dastani et al. and the ADIPOGen Consortium, is a multi-ethnic meta-analysis of adiponectin associations in nearly 46,000 subjects. Levels of the hormone adiponectin are inversely associated with type 2 diabetes and other metabolic traits. 
  • Leptin GWAS, published by Kilpeläinen et al., is a meta-analysis of 23 association studies, with over 32,000 subjects, for unadjusted and BMI-adjusted circulating levels of leptin. Levels of leptin, a hormone secreted by adipocytes, correlate with adiposity measures such as body fat mass and body fat index. 
  • Four MAGIC HbA1c GWAS ancestry-specific datasets. This study from the Meta-Analyses of Glucose and Insulin-related traits Consortium (MAGIC), published by Wheeler et al., is a meta-analysis of HbA1c levels. The results in the T2DKP comprise separate datasets for four ancestries: African American (~7,600 subjects), East Asian (~21,000 subjects), European (~124,000 subjects), and South Asian (~9,000 subjects). 
  • CHARGE Fatty Acid GWAS, published in two papers (Guan et al. and Wu et al.) from the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium, includes associations with plasma levels of nine different fatty acids and phospholipids in nearly 9,000 individuals of European ancestry. 
  • Global Urate Genetics Consortium GWAS, published by Köttgen et al., is a meta-analysis of variant associations with levels of serum urate across more than 100,000 European-ancestry individuals. Although levels of serum urate are not on their own considered to be associated with T2D risk, a recent clustering analysis (Udler et al.) found that loci associated with decreased serum urate concentrations cluster with loci associated with liver function and lipid metabolism, potentially identifying biochemical pathways involved in a sub-type of type 2 diabetes.
  • JDRF Diabetic Nephropathy Collaborative Research Initiative GWAS (Salem et al.) represents pre-publication sharing of results from a genome-wide association study across nearly 19,500 individuals with type 1 diabetes who were assessed for a wide range of phenotypes related to kidney function and kidney disease. Results for associations with ten different phenotypes (both unadjusted, and adjusted for HbA1c and BMI) are integrated into the T2DKP.

Summary statistics for all of the datasets listed above except for the unpublished JDRF DNCRI dataset are available for public download. In addition, researchers in the Slim Initiative in Genomic Medicine for the Americas (SIGMA) Consortium have made three sets of summary statistics available for download: GWAS SIGMA, SIGMA exome chip analysis, and the exome sequences from the SIGMA cohorts that are included in the 19k exome chip analysis dataset. Find details and the download links here.

This release also includes a new feature: a column showing the gene in which a variant resides is now included by default in the table of High-impact variants on the Gene page.

The High-impact variants table on the CDC123 gene page

This table is meant to highlight variants that impact the coding sequence of a gene and thus may exert their effects on disease risk through that gene product. However, since the Gene page integrates information about variants across the region of a gene, including 100kb of up- and downstream sequences, the variants shown in this table may be located in the coding sequences of nearby genes rather than in the gene that is the focus of the Gene page. In the example shown above, from the CDC123 gene page, only one of the variants in the table affects the CDC123 coding sequence; the inclusion of the Gene column makes this immediately clear.

As 2018 draws to a close, we on the T2DKP team extend our best wishes to T2DKP visitors for health, happiness, and good research results. We look forward to hearing from you in the New Year!