PSB 2016

Despite the increasing prevalence of Whole Genome Sequencing (WGS) and Whole Exome Sequencing (WES) in clinical settings, it is still very difficult to determine the causal variant for any given disease and most such explorations fall into research contexts rather than routine clinical diagnostics. To truly realize precision medicine, we must embrace the idea that all diseases are rare in that each person has their own diversity of genotypic, phenotypic, and environmental variation.

This session will explore problems in combining genotype and phenotype data to support rare disease and/or precision diagnostics. This includes the use of large-scale genomic and related data, such as SNPs, population frequency, structural variants, epigenetics, RNAseq and other expression data, interactome data, and a variety of pathogenicity measures. This would also include phenotypic analysis performed by EHR text mining, use of clinical terminologies such as SNOMED-CT and the Human Phenotype Ontology, patient self-phenotyping, rare disease registry data, lab tests, and chronological change. We particularly welcome submissions that span many types of these data. The focus will be on methods for determining causality and clinical actionability, evidence evaluation, and comparisons across approaches.

Session Topics

We welcome submissions about: (1) Ideas on how to address comparisons of approaches, (2) Improvement of emerging solutions, (3) adaptation of rare disease genotype-to- phenotype methods in the context of personalized medicine for more common diseases, and (4) results from such methods including validated diagnoses based upon integration of data across diverse genotype, phenotype, and species sources. Examples of topics and problems within the scope of this session:

• Methods for integrating genomic algorithms with phenotypic matching algorithms
• Modeling interactions between phenotypes and environment
• Models that include pedigree data (e.g. data from family members)
• Novel methodologies for prioritizing low frequency variants found via WGS or WES
• Evaluation and comparison of existing methods and/or development of gold standard datasets to do so
• Integration of evolutionary and/or non-traditional animal genotype-phenotype association
• Semantic inference techniques for genotype, phenotype, or environmental considerations
• NLP and other feature extraction methods for curation of variant identification, mapping, and association to phenotype and/or environmental factors
• Application of somatic variation approaches towards germline analyses
• Phenotypic outcomes of signaling network perturbation
• Computational databases, software, and other resources for exploration and evaluation of genotype-to-phenotype associations
• Use of publically available high-throughput datasets (mod/ENCODE, etc.) towards rare disease studies

Session Chairs

Melissa Haendel

Associate Professor
Oregon Health Sciences University

Nicole Washington

Research Scientist
Lawrence Berkeley National Lab

Maricel Kann

Associate Professor
University of Maryland, Baltimore County