FEB 24, 2016 02:00 PM PST

Why we should clinically classify genotypes, not variants

  • Senior Director, Scientific Engagement & Public Outreach, New York Genome Center
      Nathan Pearson is a Senior Director of Scientific Engagement & Public Outreach at NY Genome Center, where he and his colleagues build simple, engaging tools that compare one whole human genome to the next, in order to cast new light on human health and history. Fascinated by human genetic diversity since working as a young student in the laboratory of Luca Cavalli-Sforza (Stanford University), he traveled with Spencer Wells (now head of the Genographic Project) to collect DNA from people living throughout Eurasia, to reveal traces of ancient migrations recorded in their genomes. In doctoral work at the University of Chicago, he studied patterns of sequence variation on the great ape sex chromosomes, spotlighting how mutation, recombination, and natural selection have jointly made the X and Y ever more different from each other. Later, in post-doctoral work at the University of Michigan, he first delved into the emerging world of high-throughput sequencing, grasping its power to address important open questions in biology. Today, working with a talented team of scientists and programmers, he helps researchers and consumers sift through the billions of letters in each human genome, and understand which important spellings make each person unique.


    A century and a half after we first probed heritability, we risk forgetting one of Mendel's own basic findings, in rushing to broaden clinical genomics to lifelong care for all. Embracing that key insight now -- as we long have in tracking the chemistry of health -- can help gird our health infrastructure for the long haul.

    Learning objectives

    • Conventionally, we clinically classify and report genomic variants.
    • That habit -- rooted in early clinical genomics (reading few genes, in few people, mostly sick) -- presumes that disease cases trace mainly to strongly acting single variants...so sweeps widespread real impenetrance, for major diseases, under a vast rug.
    • Classifying and reporting genotypes would more neatly and realistically track non-additive genomic interactions mined from emerging data, girding healthcare for broad longterm genomic utility.
    •  The history of toxicology (and of early genetics too!) starkly highlights this lesson, and suggests simple but urgent steps that stakeholders (genomicists, informaticians, caregivers, and patients) should take now.


    Show Resources
    Loading Comments...