Learning models for predicting lab values from genetic data
Principal Investigator: Dr Vladimir Vacic
Approved Research ID: 54124
Approval date: January 15th 2020
We will identify individuals who have had blood and urine tests as part of the enrolment in the UK Biobank. We will examine what have they been diagnosed with and which medication has been prescribed to them. We will look into how has medication affected their blood and urine tests and will try to correct for the effects of medication. We will examine the relationships between different blood and urine tests, as well as the relationships between the genetics that is affecting blood and urine tests. These results will allow us to design optimal models to predict the results of blood and urine tests. Such models are broadly useful for understanding how genetics affects human diseases.