Peter Robinson from the Jackson Laboratory for Genomic Medicine is the speaker at our departmental seminar series on September 19, 2017 at 11:30am‐12:30pm in TAC Auditorium (N107). Please see below his abstract and CV.
As you may know, Peter’s group developed the Human Phenotype Ontology (HPO), which is now an international standard for computation over human disease that is used by the Sanger Institute, several NIH-funded groups including the Undiagnosed Diseases Program, Genome Canada, the rare diseases section of the UK’s 100,000 Genomes Project, and many others. The group develops algorithms and software for the analysis of exome and genome sequences and has used whole-exome sequencing and other methods to identify a number of novel disease genes, including CA8, PIGV, PIGO, PGAP3, IL-21R, PIGT, and PGAP2.
– –
Human Phenotype Ontology-Driven Prioritisation of Coding and Noncoding Variants in Exome and Genome Sequencing
Peter N Robinson
In this talk, I will introduce the Human Phenotype Ontology (HPO), and present a method we developed to exploit phenotype data to identify promising candidate genes in whole exome/genome sequencing (WES/WGS) studies. Our methods filter and evaluate WES/WGS variants according to rarity, mode of inheritance, and predicted pathogenicity, and ranks the associated candidate genes according to phenotypic similarity. The latter is calculated with respect to querying human, mouse, and zebrafish mutant phenotypes using cross species phenotype mapping. Our software is being used in the National Institutes of Health
Undiagnosed Diseases Program (UDP), the 100,000 Genomes Project to evaluate WES/WGS data; I will present recent results. I will conclude the talk by presenting the Genomiser, a method we have developed for regulatory Mendelian mutations found in UTRs, promoter, enhancers, and RNA genes. Simulations using 1000 Genomes VCF files show an ability to rank the seeded causative variant in first place over an entire genome in over 60% of cases.