Charlie Kim, PhD, Verily’s head of molecular sciences

Verily—Google’s sister company focused on life sciences— has made a pair of key moves over the past year to position itself as a leader in “precision health,” which it distinguishes from precision medicine by defining it as applying data from clinical and nonclinical sources to arrive at the best intervention for a person or community.

Last September, Verily announced receiving a $1-billion investment to fund unspecified “business strategies that are additive and complementary to its current life sciences portfolio.” The investment came from a syndicate led by Silver Lake, with participation by Ontario Teachers’ Pension Plan and other unnamed global investment management firms.

Four months later, in January 2023, and days after Stephen Gillett succeeded Verily founder Andrew Conrad, PhD, as CEO, the company launched “One Verily,” a restructuring of operations that included eliminating approximately 240 jobs (15% workforce), based on the more than 1,600 employees the company employed at the time. Gillett has, however, made a few senior executive appointments, including Utpal Koppikar as CFO and Laura O’Donnell as general counsel.

Launched in 2015 as Google Life Sciences, Verily gained attention for pursuing long-range, research-heavy projects such as ”smart” contact lenses to track glucose data in diabetic patients, a collaboration with Alcon (then owned by Novartis) that ended in 2018. More recently, Verily pursued—and has since ended—work on early-stage projects such as microneedles for drug delivery, remote patient monitoring for heart failure, and its Verily Value Suite clinical analytics tool.

Verily now commits itself instead to “bring the promise of precision health to everyone, every day.” To that end, the company will continue collaborations with biopharmas ranging from Sosei Heptares (identifying GPCR targets for immune-mediated diseases) to Otsuka Pharmaceutical (developing innovative approaches to clinical trials studying treatments for adults with major depressive disorder).

On Monday, Verily joined L’Oréal to launch a multi-year study designed to help researchers better understand the biological, clinical, and environmental factors that contribute to skin and hair health over time. My Skin & Hair Journey will use a registry set to enroll thousands of self-identifying U.S. women, with the aim of developing an unprecedented body of knowledge about perceptions and experiences of skin and hair health, including the role of lifestyle, environment, wellness, and personal beauty routines. The companies announced plans to advance skincare collaboratively in January, through a partnership of undisclosed value.

Verily has also taken full control of subsidiaries that include Granular Insurance, developer of a data analytics solution designed to help self-funded employers contain healthcare costs; and the Onduo digital health platform for chronic disease management.

Verily CEO Stephen Gillett

“Our path forward is driven by the need to make deliberate choices about where we can have the greatest impact in precision health and to put ourselves on a faster path to sustained commercial success,” Gillett said in a letter to Verily employees.

It’s hard to quantify Verily’s financial success to date as Alphabet does not break out Verily results, instead including them in its “Other Bets” operations category, which combines results from several businesses “that are not individually material.” During the first quarter, Other Bets’ operating loss grew 47% year-over-year from $835 million to $1.225 billion, while revenue fell 35% compared with Q1 2022, from $440 million to $288 million, primarily generated from the sale of health technology and internet services.

“We will center our strategy [emphasis in original] on closing the gap between research and care, and becoming the data and evidence backbone for precision health, with a focus on applying AI and data science to accelerate learning and improve outcomes. We will prioritize products that accelerate evidence generation and deliver more precise interventions, and that combine solutions from both research and care to drive even greater value over time, all underpinned by Verily’s commitment to transparency, privacy, and health access,” Gillett added.

Charlie Kim, PhD, Verily’s head of molecular sciences, recently discussed the company’s approach to precision health and several recent initiatives with GEN Edge. (This interview has been lightly edited for length and clarity.)

GEN Edge: One of Verily’s business areas, especially following the company’s restructuring last January, is precision health. What is Verily’s approach to precision health and why?

Charlie Kim: Our mission is to bring the promise of precision health to everyone, every day. And when we think about what precision health is, a lot of us know about diagnostics and bringing more personalization to how we manage our care.

But increasingly, we see that a lot of the information about our health comes from outside of the traditional medical system. So increasingly there are, for example, consumer tools like wearable watches and even molecular tests you can order on Amazon or directly from diagnostic labs. And it makes sense; we’re all invested in managing our own health.

With these direct-to-consumer tools out there, this information is just accumulating at an incredible rate, but it’s not synthesized in a great way. A lot of times, the healthcare data lives in silos. Our healthcare data is sitting in our electronic health records. Data from our watches are sitting on other servers, and things like this. A lot of what we believe in is figuring out how to unlock these silos and making sure that we’re really capturing an individual’s journey throughout not just the healthcare system, but throughout their lives.

GEN Edge: How does Verily go about unlocking those silos?

Kim: You’ll hear more about that over the coming months and years. One of the things that we are really working on is putting forward our patient-facing technology.

GEN Edge: One of Verily’s patient-facing technology tools is the Onduo digital health platform for chronic disease management. How does that enhance precision health?

Kim: Onduo is a program that has an app and other tools associated with it that are already in the care setting. We deliver diabetes care to patients who are on this platform. And that is a really interesting starting point for making sure that you have the consent of an individual to use their data. By consent, I mean not just talking about having a signature on a line, but really about engaging with the patient in a way that they understand what they’re consenting to, how their data is used, being able to withdraw consent at any time, and making sure that the patient is really in control of that aspect of their journey.

That’s a really great starting point that we’re excited about. I think it’s something that’s super important. Our leadership thinks it’s super important. A lot of this comes from Amy Abernathy [MD, PhD], who is our president of product, and has come to us from the FDA to build a lot of these solutions in a way that both protects the patients, but then also can be more regulatory body facing, to make sure that the data we collect is of sufficient quality to support things like regulatory submissions.

GEN Edge: How did Verily’s restructuring enhance its precision health effort?

Kim: That was all a strategic decision to really focus on some of the core areas where we saw what we refer to as mutually reinforcing solutions. A lot of the historical Verily was sort of fit-for-purpose solutions. Now we’re moving toward building more end-to-end solutions that can solve bigger problems.

GEN Edge: How does Verily focus on its three cores of research as it has defined them: research, care, and precision risk?

Kim: Precision risk is about managing the financial risk associated with insurance. The other areas where we’re really focusing really is bringing research and care closer together. And the example I gave of Onduo is a great example, where if you’re direct-to-patient in a care setting, it starts with asking the patient if you can contact them for research opportunities. That’s the most basic form of consent there is. And that opens the door to the possibility of doing so much more, if the patients are willing.

GEN Edge: How willing are patients to consent to research opportunities?

Kim: Well, we’ll be launching a lot of this over the year, so we’ll find out.

GEN Edge: What is Verily’s research core in light of the restructuring?

Kim: On the research side, there’s everything from discovery through running trials, and even post-market. We have solutions in all of those spaces. There’s an entire backbone including registry and clinical trial management software, and clinical trial software—there’s a slight distinction between those things.

For example, our clinical trial management software can do an acquisition of signal path, and this is software that helps clinical trial sites with their operations. That’s also right at the forefront of right where the clinicians are interacting with the patients. Then, our clinical trial software is something that’s a little bit more on the pharma side, which can help bridge and make sure that the data being managed, and the patients being enrolled are technology enabled, to ensure that the trial is as sufficient and effective as possible.

The group I represent is called Precision Measures, and this is a big part of where we think research is going. And by precision measures, really, we’re thinking about, how do we get even increasingly high-resolution data.

And by resolution, that can mean a variety of different things. For example, our wearable devices, the watches we build, these can be more continuous resolution, measuring things like activity that is collected relatively passively, or through medication tracking, for which we have a very easy recording mechanism with two buttons to press and you log in the medication. We think this is going to result in higher quality data, more accurate data, with reduced patient burden in the context of a clinical trial, for example.

GEN Edge: Why? Because it’s easier to hit the buttons?

Kim: Because no smartphone is necessary, not even Wi-Fi. It’s just two buttons. You don’t have to open anything. It’s just right there on your wrist. So, we think that’s an example of how you’re going to get better data.

The programs I lead in the molecular space are pathology and also our immune profiler work. We think that genetics is going to be increasingly important as we continue to develop a better understanding of, for example, how to stratify patients for clinical trials, to understand how we can understand disease mechanisms associated with genetic risk factors, and to develop targets based on that knowledge.

On the pathology side, we have traditional pathology, where you look at an image, a stained slide, and you might get a diagnosis or a single grade out of it. But of course, this is such a rich substrate. We’re learning so much more. There’s a lot of movement across the industry, especially using AI and ML to extract more information from every section. And the more we do that, the better diagnostics we’ll develop. The faster we’ll be able to develop drugs. We’ll be able to stratify patients more effectively. We’re just scratching the surface at this point; it’s in its infancy.

GEN Edge:  So many companies talk about AI and machine learning. What’s Verily’s approach to AI? What is Verily looking for from the technology?

Kim: We’re at the forefront of some of the most novel AI advances, being in the same umbrella as Google. Some of the latest developments are around these large language models.

Our cousins over at Google recently released Med-PaLM, which is a large language model. Instead of being trained on general text unavailable on the Internet, it’s specifically trained on medical text. We believe that the purpose of it being trained on medical text is to deliver much more accurate medical guidance and information back to anyone running a query.

And it’s still very early. I would say all of these generative AI approaches using these large language models are still trying to figure out the right use cases where the tuning for accuracy and confidence and things like that are going to be really important. In the healthcare space, it’s super important that all of this is absolutely spot-on in terms of accuracy, so there’s still work still to be done there.

But the future we envision is that these sorts of tools are going to be patient-facing, physician-facing. It’s going to be pervasive through a lot of the activities that are very manual right now. Everybody should win by getting the most accurate information possible, if we do all of this right and develop the tools in the right way.

GEN Edge: What’s driving this adoption? Is it healthcare providers shifting? Is it biopharmas?

Kim: I think it’s everyone! There are applications for AI across the preclinical through clinical through care delivery phases. I think we’re going to see all sorts of interesting applications. The ones that you’re seeing in the media right now, of course, are the ones that are very high visibility and tangible to all of us. But I think it’s going to be pervasive in everything starting from just managing data quality and monitoring data quality. There are going to be all sorts of innovations that will be invisible to us, and that will be ubiquitous.

GEN Edge: Verily has identified three essentials for precision health: AI, which we’ve been discussing; data, and health technology. In terms of data and health technology, how does Verily generate the data it needs to develop interventions?

Kim: That is such a tricky question because we are across so many different therapeutic areas with such different approaches, everything from an intervention guided by, for example, a watch, versus the programs I lead in immunology, where we are working to prioritize drug targets. The answer is nuanced.

GEN Edge: How has Verily applied its precision health approach in immunology?

Kim: One of the things that we’ve really been interested in the immunology space is what the industry has come to know as genetics-driven drug discovery. Basically, this concept is that you use the power of evolutionary selection and the identification of genetic risk factors for disease through GWAS (genome-wide association studies) as a starting point for looking into what disease mechanisms might be causal downstream of those genetic risk factors.

The tricky part is figuring out what those pathways are that are perturbed. Some of the work we’re doing with our immune profiler platform is really understanding the functional changes associated with genetic risk factors, and using that in combination with a lab platform we have where we generate very high-resolution functional genomic data, associate that with genetic information, and using purpose-built analytical pipelines prioritizing drug targets, to identify those that have the strongest genetic and functional genomic evidence.

We’ve invested heavily in this because the literature is indicating that drug targets that have this kind of evidence have a 2X greater chance of reaching FDA approval. That translates to over $500 million saved on every single drug launched. It would be transformative if we could fully achieve that kind of success in terms of drug development today.

GEN Edge: Can you elaborate on the target selection process by which Verily helps identify targets?

Kim: A great example of this is our collaboration with Sosei Heptares, which is one of the world leaders in structure-based drug design for G protein-coupled receptors (GPCRs), the most successful class of drug targets in history. In part, GPCRs have been so successful because they’re a relatively large class of molecules. There are about 800 GPCRs encoded in the human genome. About half of them are at least annotated as a full factory. Of the remaining 400, only about 100 of them have an associated approved therapeutic.

When we began talking to Sosei, we were able to articulate the challenge of these 300 undrugged GPCRs: Which ones could they go after? Could we use some of our data and analytics to help them prioritize which ones to go after? We’re about a year and a half into the collaboration. We’ve worked with them to leverage our proprietary database and analytical pipelines to narrow down that list of 300, all the way down to just a few tractable targets with the strongest genetic and functional genomic evidence.

We’ve taken those into the labs and validated them, and discovered some novel biology associated with each of those targets, and are now ready to move them into Sosei’s sweet spot of structure-based drug design. These are some very novel targets, some with some established understanding of what they do, and it’s a nice mixture of both.

GEN Edge: Is the data collected by Verily or partners? Healthcare providers? Drug discoverers and the like? Is it their data that Verily is synthesizing, or is Verily collecting data?

Kim: It’s a combination of everything. Data is everywhere. We know we can’t solve all the problems by ourselves—no institution can. Our business model is heavily reliant on partnering and finding complementarity with our partners, so that we can do a lot more together than separately.

GEN Edge: Where is precision health headed?

Kim: Obviously we’re not alone in pushing forward these frontiers. All kinds of interesting novel devices are being developed that are collecting physiometric data in all sorts of interesting ways that previously we didn’t have access to. This includes things like lots of people doing sleep monitoring, for example; the quality and quantity of sleep we get is something that’s closely tied to our wellness. All of these things are moving forward as data collection exercises and insight generation.

But the big challenge we’re going to face is: how do we stitch it all together into meaningful insights that drive behavior change for all of us? That’s where we have to keep our eye and we’re going to see a lot of synthesis of these right now fairly disparate activities coming together into real change in terms of the information, the actionable insights we get back.

Previous articleWGS More Effective than Targeted Gene Sequencing For Diagnosing Genetic Diseases in Newborns
Next articleBoston Calling: Genomenon Acquires Boston Genetics for Rapid Human Genome Curation