WP leaders: Tim Frayling and Giraj Chandak
The genomics component of the GIFTS project has been designed to understand the influence of maternal genetic variants on foetal phenotype and markers of early growth, as well as the effect of maternal variants on the long-term risk of obesity & adverse metabolic outcomes in children. The direct effect of foetal gene variants on body constitution and related phenotypes is also being explored. Finally the interaction between genetic variants and different interventions, and their influence on the risk of intermediate traits related to metabolic syndrome will be studied. DNA samples were collected from South Asian populations residing in India, Bangladesh and the U.K. All the data used in WP6 were generated by single nucleotide polymorphism (SNP) genotyping arrays (Illumina human exome beadchip), The chip includes 250,000 tag SNPs and rare variants and thus allows exploration of the role of both coding and non-coding markers.
Data generation and analysis was divided into two arms due to an embargo on the shipment of DNA samples outside India and the large number of samples: The samples from India were processed and analysed by Dr Giriraj Chandak and his team at the Centre for Cellular and Molecular Biology (CSIR-CCMB), a CSIR Institute in Hyderabad, India. The remaining samples were to be processed at QMUL (DNA preparation) and by Prof Tim Frayling at the University of Exeter, UK for genotyping. Following analysis at each centre according to a common protocol, meta-analysis was planned using the combined data.
Data generation and analysis – India
After quality control checks on the array data, 1325 samples (678 children and 647 mothers, including 627 mother child pairs) were available for further analysis. Unbiased association analysis of variants with various traits was performed to try and identify novel variants. We also carried out targeted association analysis of established variants with various traits to establish the role of identified variants.
Finally, we performed Mendelian Randomization (MR) analysis:
– Using individual variants (established and novel, if any).
– Using genetic risk scores based on the number of effect alleles.
– Using weighted risk scores based on the effect size and the number of effect alleles.
Using the above approaches we have identified several loci that may indicate causal influence of these variants on determining birth weight. Analyses on other traits using similar approaches are underway.
Data generation and analysis – UK
In this arm of the study DNA was extracted from blood samples collected from WP2 Bangladesh, WP3 Bangladesh (intervention trial), WP2 London and the FEATURE study. A total of 1800 samples including 38 replicates have been prepared for processing on the Illumina Human Core Exome Beadchip and for epigenetic studies.
Unfortunately, we have not been able to array the samples on the beadchip for genotyping due to a number of confounding factors. We plan to run these samples using a new Illumina GWAS chip in the near future that will enable the meta-analysis of data sets.