oa Large Scale Omics Analysis of Type 2 Diabetes in the Qatari population

Background: The prevalence of Type 2 Diabetes (T2D) in Qataris has been recently recorded (as to the Qatar Bio Bank) to be around 14-15%. T2D associated macro- and micro-vascular complications leading to retinopathy, neuropathy, nephropathy, as well as cardiovascular complications are among the most known factors leading to death. Both genetics and environmental/life style factors play roles in the pathophysiology of T2D. The advent of the Qatar Genome Project and whole genome sequencing of the Qataris will provide a wealth of genomic information on several diseases inherent in the Qataris and specifically high prevalent ones as T2D and its complications. Profiling other “omics” data as gene expression, metabolomics, proteomics, and epigenetics among others became imperative for unraveling the complex nature of T2D and the effect of both genetic and environmental factors. Objectives: In this study we focus on integrating metabolomics and genomics data for identifying their associations to T2D using nearly 1000 Qatari samples. We achieve that in the following main steps: 1-Identifying associations between metabolites and exome variants (metabolic Quantitative Trait Loci (mQTL)) in 1000 Qataris for the sake of identifying mQTLs (metabolic Quantitative trait loci) linked to both common and rare variants in Qataris. 2-Identifying T2D associated metabolites and the pathways enriched in those metabolites. 3- Investigating the associations of T2D to genes/exome variants linked to T2D through the identified mQTLs (step 1). Materials and Methods: Genomics Data: 1000 samples were collected from Qatari individuals (50% with T2D), where DNA was extracted and used for both whole exome sequencing (n = 614) and genotype arrays (n = 382)(4 samples were removed after quality control, leading to 996 samples in total). Exome and array data were imputed after being filtered (MAF> = 0.05, pHWE>10-6, genotype call rate > = 98%) based on phased 108 Qatari whole genomes as a reference panel, using shapeit and Impute2 software packages. A total of 1.6 million variants from the imputed exome were used for discovery analysis of the mQTLs and the array data was used for replication. Metabolomics Data: Serum samples for the 1000 samples were used for profiling metabolomics using a recent advanced Metabolon platform (DiscoveryHD4). This platform utilized a Waters ACQUITY ultra-performance liquid chromatography (UPLC) and a Thermo Scientific Q-Exactive high resolution/accurate mass spectrometer interfaced with a heated electrospray ionization (HESI-II) source and Orbitrap mass analyzer operated at 35,000 mass resolution. A total of 1303 metabolites were measured on that platform, and we were only left with a total of 826 metabolites (including 249 unknown metabolites) for the analysis after quality control (missing values <  20%, and outliers removed). Association Analysis: Linear regression models were used for computing associations between Metabolites and SNPs, as well as between metabolites and T2D after correcting for covariates. Results For metabolomics – Exome wide associations [Yousri et. al 2017], we discovered and replicated 21 unique mQTLs associated with common variants (MAF > 5%) and discovered another 12 mQTLs associated with rare variants using burden tests and single variant analysis. Overall, 45% of the discovered loci are novel ones. We also replicated 19% of the known mQTLs in European populations using the discovery cohort. Regarding metabolomics of T2D, we identified 190 metabolites associated with T2D, spanning pathways of fatty acid metabolism, phospholipids, sphingolipids, phenylalanine and tyrosine metabolism among others, and where 40% of the metabolites were newly identified (new to our previously reported T2D metabolites in different biofluids [Yousri et al 2015]). We also identified several associations between genes known to be associated with T2D and T2D associated metabolites. In summary, we have both identified mQTLs from large scale metabolomics (m = 826) and whole exome sequencing, and identified T2D metabolites using a large sample set (∼1000 samples), and used those to reveal the links between an intermediate phenotype – metabolite - and the genes/variants known to be associated with T2D. This is considered the first time this study is done in the Qataris integrating omics data on a large scale. References: [Yousri et. al 2017] Noha A. Yousri, Khalid A. Fakhro, Amal Robay, Juan L. Rodriguez-Flores, Robert P. Mohney, Hassina Zeriri, Tala Odeh, Sara Abdul Kader, Eiman Aldous,Gaurav Thareja, Manish Kumar, Alya Al-Shakaki, Omar M. Chidiac,Yasmin Mahmoud, Jason G. Mezey, Joel Malek,Ronald G. Crystal5, Karsten Suhre. “Whole Exome Sequencing identifies common and rare variant Metabolic Quantitative Trait Loci in a Middle Eastern Population”. Accepted in Nature Communications. [Yousri et al 2015] Noha A. Yousri, Dennis O. Mook-Kanamori, Mohammed M. El-Din Selim, Ahmed H. Takiddin, Hala Al-Homsi, Khoulood A.S. Al-Mahmoud, Edward D. Karoly, Jan Krumsiek, Kieu Thinh Do, Ulrich Neumaier, Marjonneke J. Mook-Kanamori, Jillian Rowe, Omar M. Chidiac, Cindy McKeon, Wadha A. Al Muftah, Sara Abdul Kader, Gabi Kastenmüller, Karsten Suhre, “A systems view of Type 2 Diabetes-associated metabolic perturbations in saliva, blood and urine at different time-scales of glycemic control”, Diabeteologia, August 2015.