Skip to navigation Skip to main content Skip to footer

Approved research

Use of large scale, real world genotype ?to-phenotype data to identify disease sub-types that will enable precision medicine drug discovery and development

Principal Investigator: Dr Melissa Miller
Approved Research ID: 37797
Approval date: February 21st 2018

Lay summary

a.Perform phewas on phenotype-associated loci/variants to elucidate risk/benefit profiles for coincident diseases. b.Perform phewas on targets of known drugs and public compound libraries to identify repurposing opportunities and probes for validation strategies. c.Explore the impact of correlation structure in high-dimensional phenotype-genotype data on multiplicity adjustments and testing procedures. d.Develop and validate electronic phenotyping algorithms for patient subtyping and disease progression (e.g. heart failure) using statistical, machine learning and NLP approaches on biochemical, imaging, exposure, text record and coded data. e.Phenotypic Adjacencies: Identify disease co-occurrence and co-morbidities for diseases (e.g. PD-related depression, T2D-related pain) and explore outcomes in a longitudinal setting. These objectives support (i) dissecting the genetic underpinnings of disease and clinically meaningful phenotypes, (ii) refining therapeutic targets by assessing benefit/risk profiles in human data, (iii) accelerating target validation through chemical probe characterization, and (iv) gaining better understanding of patient profiles for monitoring, disease progression and subtyping. We will explore associations of both phenotype-associated loci and genes of known safe drug targets to identify novel-disease phenotypes, potential mechanisms, or repurposing opportunities. We will use multiple types of data (self-reported data, clinical data, EMR data, biomarkers) to define novel algorithms to define complex diseases and will perform GWAS to identify genetic underpinnings of these diseases. We will use these same types of data and developed phenotype algorithms to define sub-types of disease, rapid disease progressors, and to identify novel disease co-morbidities or occurrences. We would like to maximize the sample size and request access to the full cohort.