Mainstream news stories often mistake technology for science. Such conflation likely reflects how gizmos and knowledge, like arteries and veins, form a virtuous cycle: understanding nature requires, and refines, good tools.
In genomics, that symbiosis burns brightest at Advances in Genome Biology & Technology (AGBT), the yearly nucleic acid trip in Florida, where new sequencing boxes jostle alongside key findings from the current machines. While hardware dominates the AGBT buzz – think Oscar intrigue, but for biogeeks – the science on tap highlights how those technical marvels address big biomedical questions. Look for several key themes this year at the conference, which runs Feb. 20-23.
Beyond low-hanging fruit
Genomicists have long known that our genomes abound with rare but important variants, and that exons, the easily studied genome segments that encode proteins, are just the tips of an iceberg of functionally important DNA. But only lately have we started to thoroughly catalog those rare variants, and the genome segments that govern how much of each protein gets made, when, in what tissues, and under what conditions.
Troves of data from sequencing big population samples and surveying gene regulation under varied conditions have been the talk of human genomics. Expect AGBT speakers, in rafts of talks on gene expression and disease, to drool over how these new public data can drive biomedical research, and to propose concrete plans to that end.
For genome interpreters, the new findings on rare variants and regulatory segments of the genome are both boon and challenge: they can clarify why people get sick, but only if we refine statistical methods for spotting patterns in vast, complex data. Luckily, we’re already seeing such methods flag how rare variants cluster in particular genome segments – exons or otherwise – in people with the same disease.
Towards fast, reliable clinical whole genomes
While healthcare is the consensus killer (lifesaver?…) app for whole genome sequencing (WGS), the method is still slower and costlier than targeted sequencing – and, today, may miss key variants or, conversely, yield too much information, posing dilemmas on interpreting and reporting findings.
Folks at AGBT thus think a lot about making WGS fast, cheap, and diagnostically reliable. That starts with lab and computer methods to more accurately characterize genomes: look for claims that platform X, sample prep trick Y, or program Z pieces together especially long snippets of a genome, revealing which variants occupy the same chromosome copy, and which chunks of that chromosome may have been cut, copied or moved elsewhere.
But also look for streamlined pipelines to do all this—and interpret the sequenced genome – faster than ever. Getting the right answer fast is key for clinical labs, who want to scale up WGS via auditable, repeatable workflows that extend lives, meet regulations, and cut costs.
To do so, labs need more than just an accurate snapshot of a patient’s genome; they also need to know which of the patient’s variants are found in other sick or healthy people, and which of the resulting leads most plausibly underlie the patient’s disease.
Thus expect clinical genomics speakers to discuss community efforts to a) classify variants by consensus criteria that let interpreters quickly triage them when found in patients, and b) help labs share genomic findings from their patients in productive ways, so that colleagues in Denver, Delhi, and Dubai can quickly leverage insights from across the planet to solve local cases of rare disease.
But such sharing will require new policies, including smarter informed consent protocols that safeguard patients’ health and privacy, but let humanity benefit from gathered data. Of the >6 billion DNA letters in a genome, very few can directly inform a person’s healthcare today—but the rest of those letters can help build a critical mass of knowledge on genetic variation that benefits everyone. As such, secure and productive sharing of genome and associated phenotype data has become a key topic at AGBT and similar meetings.
From one genome to many
Many AGBT talks will highlight how fast, cheap, and sensitive sequencing technology lets us peer deeper into the diversity of genomes around us – and, increasingly, within us.
Over the past decade, we’ve moved beyond thinking monolithically of ‘the’ human genome, recognizing that only by surveying how genomes vary do we learn how they shape our traits. Today, thousands of people’s genomes have been comprehensively sequenced, and much of our field’s buzz centers on new methods for making sense of the resulting data. One front worth watching is how we’re adapting methods for studying small families to the challenge of comparing genomes in huge pedigrees and other population samples, to assess kinship among thousands of people at once and, in so doing, help better understand disease heritability.
But just as sequencing has led us to think of our genomes at the population scale, it now points to the many genomes that each of us harbors. That’s clear first for tumors: clumps of our own cells whose genomes are distinctive in ways that let them divide dangerously fast. Many AGBT talks will look at those genomes—including how they vary within a tumor, and over time.
But other talks, likewise leveraging sensitive sequencing methods, will highlight how our internal genetic diversity extends further: to eggs and sperm, whose diversity carries on into future generations; to immune response cells, whose genomes programmatically rearrange; and to the foreign cells within us, including our thriving microbiomes, as well as cell lines derived directly from our mothers or children during gestation, which can, remarkably, live on for decades.
All these cells may play key roles—both positive and negative—in health. And, thanks to an embarrassment of technical riches, geneticists are starting to characterize their diversity with stunning precision. Look for early results next week, and ever deeper insights at AGBTs to come…