Skip to navigation Skip to main content Skip to footer

Approved research

Identifying genetic tissue-specific disease risk of type 1 diabetes

Principal Investigator: Mr Daniel Ho
Approved Research ID: 51306
Approval date: September 12th 2019

Lay summary

The organization of the DNA within each of your cells is the end result of all the processes (e.g. transcription, repair, and replication) that are occurring within the cell or tissue. This means that we can potentially integrate information on the 3-dimensional organization, of DNA, with patterns of genetic variation, transcription, and phenotype measures to improve our understanding of the genetic basis of complex phenotypes. In fact, this approach has been shown to be useful recently. We will use a form of machine learning to combine data on the genetic variants that are associated with type 1 diabetes, gene expression data, and data on the 3D structure of DNA in immortal cells to identify the biochemical pathways that are affected in type 1 diabetes. We have a reference table that allows us to identify which of the genetic changes associated with type 1 diabetes affect which genes. We will use this reference table to convert an individual person's profile of genetic variation (their genotype) into a score that 1) predicts their risk of developing type 1 diabetes and 2) identifies which of their tissues make the greatest contribution to this risk. We will train our machine learning algorithm on the UK Biobank data to dissect the biospecimens that are held in this repository. We will then validate the results with the Wellcome Trust Case and Control Consortium data. This study requires the individual genotypic and phenotypic data (e.g. diabetes medical conditions) from the UK Biobank to maximize the predictive power of our machine learning risk predictor. Preliminary analyses have demonstrated that this approach has the potential to work and make a significant contribution to our understanding of type 1 diabetes. As the result of the study, we will improve our ability to predict a person's chances of developing type 1 diabetes and identify potential drug targets that may reduce type 1 diabetes risk in children and adults.