How we can accelerate the translation of genetic findings into benefits for patients

As the new academic year — Broad’s 20th! — gets underway, it’s a good time to reflect on what we’ve achieved and more importantly, what lies ahead.

Science has progressed remarkably over the past 20 years, with advances in human genetics being particularly noteworthy. When the Broad Institute launched, the sequence of the first human genome had been only recently completed and, led by Broad researchers, we were well on the way to creating the HapMap — the first major international genome-scale map of human genetic variation.?

International initiatives, in many cases led by Broadies, showed that the concept of using genetic variation as a measure of disease risk was no longer just theoretical. For disease after disease, risk alleles began to emerge, with increasing statistical confidence as the scale of the studies grew. And for some disorders, severe mental illnesses in particular, genetics provided a biological foundation for the first time.

There were some important surprises along the way too. No one imagined that the cost of genome sequencing would fall as quickly and dramatically as it has, by as much as a million fold, according to some estimates. We also came to understand the power of studying the genomes of diverse populations, with the discovery of risk alleles in individuals of African or Latin American descent that were entirely missed in European ancestry populations. And, while we initially thought that low penetrance risk alleles for common disease would be useful for biological discovery but not for individual patient risk assessment, polygenic risk scores – a summation of genetic variation across the genome – have proven useful at least for some diseases such as breast cancer and cardiovascular disease, and are already being implemented clinically.

So where are we now? The search for disease-associated risk alleles is certainly not over, but we have already identified thousands of such variants — mostly in non-coding, uncharacterized regions of the genome. Our task now must be to turn that mountain of variants into a smaller number of biological hypotheses that can inform therapeutic or preventive strategies to benefit patients. But the path to converting genetic variation into mechanism-based intervention is not obvious. Figuring out how to do this must be the work of the Broad’s next chapter.

We know that increasing the scale of human genetics studies helps build statistical confidence in disease risk association, but it’s also clear that genetics alone will not be sufficient to decipher the biological consequence of genetic variation. Rather, we must extend our analyses beyond genetics to the study of the transcriptome, proteome and metabolome — ideally at single-cell resolution — and we must understand how these multi-omic profiles impact each other across cell states in healthy and diseased tissues.

This will require bold new approaches to bringing scale to the problem, while also being mindful of where scale is not the answer. We’re now comfortable with scaling human genetics, but even this is a relatively new phenomenon (when I was a postdoc, human genetics largely took the form of crude positional cloning and testing candidate genes one at a time). Today, we think of mechanistic studies as being fundamentally unscalable — such studies only being possible by focusing on one protein at a time, each being pursued by individual scientists. Yet if each of the thousands of variants must be individually worked up mechanistically, we will move far too slowly. And patients are waiting.?

Our challenge for the years ahead must therefore be to figure out how to thoughtfully bring scale to bear on mechanistic studies, while also recognizing when it’s time to use more traditional biochemical and cell biology methods to get the level of detail that is essential for drug discovery. Broad is uniquely positioned to thread this needle because of our track record, talent, and our two-decade history of figuring out precisely what scientific transformations will be needed — and then working collaboratively and creatively to lead the way there.?

Getting this right will require building new communities that span disciplines — bringing together scientists who are steeped in the world of unbiased, large-scale biology (but who might be less adept at mechanistic follow-up) with those who are more comfortable in hypothesis-driven research but are not yet accustomed to systematic, genomics-style strategies.?

One obstacle in creating these communities at the interface of fields is the lack of a shared language. We must take it on as our challenge to create such common languages — for example, at the interface of neuroscience and human genetics; at the interface of machine learning and cell biology; at the interface of functional genomics and drug discovery; and at the interface of clinical medicine and discovery biology. I believe that the most exciting and important breakthroughs in biomedicine will occur at these interfaces, and Broad is ideally positioned to take this on. It will require courage, patience, mutual respect, and above all, a drive to work together (especially with our hospital partners) to solve medicine’s most pressing problems with urgency, recognizing that these are not problems we can solve alone.?

So what will success look like? For most human diseases, we will have converted a pile of bewildering variants into a limited number of discrete therapeutic hypotheses that can be tested in experimental models and eventually in people. We will have created new experimental cell models (e.g. using scalable gene editing in iPS-derived cells) and we will have figured out how to learn from patient-derived samples (e.g. blood, tissue) beyond genomic DNA. We will have established compelling evidence that integrates human genetics, tissue biology and pre-clinical genetic manipulation to formulate therapeutic hypotheses that will give drug hunters unprecedented confidence that they’re going after the right targets. That will translate into clinical trials being more likely to yield positive results. And, success will mean that methods developed at Broad will be used around the world.?

Longtime Broadies will recognize this list as “ambitious but doable,” partly because we know how powerful this community is when we come together.

Broad is a special place, full of people who really want to invest in each other, and who want to advance our collective mission of improving human health through transformative and open science. I’m grateful to Mark Daly, the Medical and Population Genetics community at Broad, and members of Broad’s Scientific Leadership Team for taking on the challenge of figuring out how our collective aspiration of Human Genetics 2.0 might take form.

Of course our ultimate measure of success will be helping patients. Yes, many papers will be published along the way, but we should not lose sight of our mission to use the power of science to relieve suffering from disease. I can’t imagine a more inspiring goal, and I’m excited to continue our work together to make this a reality!