Skip to content

The Future of Genetic Networks in Decoding Complex Disease

On April 3, 2019, CIFAR convened a roundtable on a topic at the vanguard of genomics: how combinations of genetic variants yield phenotypes, and how knowledge of genetic interactions can be exploited to interpret genomes. Attendees came from across academia, clinical institutions and industry, including Fellows from CIFAR’s program in Genetic Networks; clinical experts from medical genetics institutes and national genomics projects; and scientific and business leaders from companies working in the genomics sector, from consumer genetic testing, bioinformatics analysis, to drug development.

Discussions at the roundtable focused on novel experimental approaches and computational methods that may improve our understanding of the functional impact and clinical utility of genetic networks. Through brief primer talks and moderated discussions, the group explored how the combination of genomic and phenotypic data obtained from ever-larger cohorts, both population-wide and disease-specific, are deployed to identify the functions of genetic variants; how novel genomic approaches and molecular tools (such as CRISPR) are being used in model organisms and cell lines to better understand genetic networks; and how advances in genome sequencing technologies and computational approaches are contributing to the development of clinically actionable genetic information. This meeting highlighted the progress in understanding how genetic variants and interactions contribute to complex human diseases, identified key areas where advances in technology and practice (such as data sharing) are still needed to advance precision medicine, and paved the way to further collaboration between researchers, clinicians and industry.

Impacted Stakeholders

  • Healthcare systems
  • Organizers of precision medicine initiatives
  • Genetic testing companies
  • Pharmaceutical companies (developing personalized / targeted therapies)
  • Developers of sequencing technologies and bioinformatics platforms

Key Insights

  • National genomics programs and genetic testing companies worldwide are gathering ever larger cohorts of both healthy individuals and patients with rare and common diseases, on the order of tens to hundreds of thousands. If collected alongside clinical / phenotypic data, such cohorts can contribute significantly to understanding biological functions of genetic variants.
  • Studying genetic networks is important, as it is possible for two individuals with the same disease to have different disease-causing variants and yet the pathways in which the variants interact are shared. An understanding of genetic networks can therefore help identify potential causes of disease or other biological functions that should be considered, pointing clinical “detectives” in the right direction. It can also help inform the selection of specific patients or subsets of patients for further study.
  • In studying gene variants, it is important to take into account gene duplication and redundancy of function, as variants in one context may do something different in a different context. Complex model organisms can provide a tool to probe which changes cause specific effects.
  • New tools for making targeted genetic changes, including genome editing techniques such as CRISPR, have made it more feasible to experimentally query the function and interaction of genetic variants.
  • Proactively creating and studying all possible variants of genes of interest may speed up the interpretation of clinical sequencing data. Simultaneously perturbing pairs or even triplets of genes are revealing previously unknown genetic interactions that affect phenotype.
  • Advances in computation, including artificial intelligence-related applications (such as natural language processing) and cloud computing, are contributing to new and faster ways of processing large amounts of genetic and clinical information, and of mapping genetic interactions and pathways.
  • One possible approach to making genome analysis more readily applicable in the clinic is to shorten the timeframe of the pipeline through optimization or automation of various steps, including sample collection, sequencing technology and data analysis.

Priorities and Next Steps

  • As more genomic data are generated, geneticists need to decide on the appropriate level(s) of analysis – whole genomes, exomes, transcriptomes, or individual candidate genes – as well as the focus(es) of their efforts – identifying all variants, elucidating the functions of known pathogenic variants, or delineating networks and pathways.
  • The research and clinical communities, public funders and consumer-driven companies need to devise policies and technical solutions for improved data sharing. Considerations should be given to allowing more researchers to collaborate on the functional characterization of genetic variants in order to drive actionable predictions, while at the same time protecting the privacy of patients or consumers who contributed the data.
  • The genomics community should come to an agreement on terminology (e.g., loss of function vs. toxic gain of function, or pathway vs. module as a unit of interaction) as this could present a barrier to understanding and progress.
  • Researchers, clinicians and technologists should collaborate on optimizing the pipeline for genomic analysis, in order to gain the greatest functional insights from genomic data and translate them to the bedside.

Roundtable Participants

  • Catherine Ball, Ancestry
  • Anastasia Baryshnikova, Calico Life Sciences
  • Serafim Batzoglou, Illumina
  • Charlie Boone, University of Toronto / CIFAR
  • David Botstein, Calico Life Sciences / CIFAR
  • Andrew Carroll, Google
  • Kevin Haas, Myriad Women’s Health
  • Dann Huh, Biogen
  • Johnny Israeli, Nvidia
  • Stephen Kingsmore, Rady Children’s Institute for Genomic Medicine
  • Bertram Koelsch, 23andMe
  • Jason Moffat, University of Toronto / CIFAR
  • Chad Myers, University of Minnesota / CIFAR
  • Michael Phillips, Sequence Bio
  • Jeffrey Reid, Regeneron Pharmaceuticals
  • Fritz Roth, University of Toronto / CIFAR
  • Richard Scott, Genomics England
  • Dayag Sheykhkarimli, University of Toronto
  • Unnur Thorsteinsdottir, deCODE genetics
  • Olga Troyanskaya, Princeton University / CIFAR
  • David Whitcomb, University of Pittsburgh

For more information, contact Amy Cook, Senior Director, Knowledge Mobilization.

Support Us

CIFAR is a registered charitable organization supported by the governments of Canada, Alberta, Ontario, and Quebec as well as foundations, individuals, corporations, and international partner organizations.

MaRS Centre, West Tower
661 University Ave., Suite 505
Toronto, ON M5G 1M1 Canada