Rafael Rosengarten of Genialis On 5 Things We Must Do To Improve Patient Outcomes for Underrepresented Populations
Build diverse teams. We want to build companies that have a diversity of opinion, a diversity of experience, and a diversity of professional contributions. As we’re building our tech companies and the companies that are going to bring solutions to healthcare, we should aim to maximize the diversity on as many axes as possible, especially diversity of experience and perspective.
Healthcare disparities continue to affect underrepresented populations, leading to unequal patient outcomes. Social, economic, and cultural barriers often hinder access to care, appropriate treatments, and equitable health services. How can we bridge these gaps and ensure that all patients, regardless of background, receive the highest standard of care? In this interview series, we are talking to healthcare providers, policy makers, community leaders, researchers, and anyone who is an authority about “How We Can Improve Patient Outcomes for Underrepresented Populations”. As a part of this series, I had the pleasure of interviewing Rafael Rosengarten.
Rafael Rosengarten is CEO and co-founder of Genialis, the company revolutionizing diagnostics and drug development with RNA and machine learning. He envisions a world in which healthcare delivers the best possible outcomes for patients, their families and communities. Before leading Genialis’ effort to realize the promise of precision medicine, he spent nearly 20 years in biomedical research, publishing in the fields of evolution, immunology, bioengineering and genetics. Dr. Rosengarten is also a board member and co-founder of the Alliance for AI in Healthcare (AAIH), a non-profit global advocacy organization. Rafael attended Dartmouth College before earning his doctorate from Yale University, where he was an NSF Graduate Research Fellow. He went on to postdoctoral training in Jay Keasling’s synthetic biology group at Lawrence Berkeley National Laboratory, Joint BioEnergy Institute (JBEI), where he co-invented the j5 DNA assembly design automation tool (which has since been commercialized by TeselaGen Biotechnology). This was followed by a National Library of Medicine fellowship in Biomedical Informatics at Baylor College of Medicine.
Thank you so much for joining us in this interview series! Before we dive into our interview, our readers would like to get to know you a bit. Can you tell us a story about what brought you to this specific career path?
I think of myself as an accidental entrepreneur. I always knew I wanted to be a scientist. At age seven, I learned to snorkel with my cell biologist uncle off the coast of Villefranche-sur-Mer, France, near his marine laboratory. From that day, I knew I wanted to study the natural world. However, I had no idea how that interest would evolve over the course of my education and academic career.
There was a time, through undergrad and during a 3 year stint working in restaurants after college, I thought I would become a professional chef. But the long late hours and grueling lifestyle were more than I wanted for a long-term career. I went back to grad school and received my PhD.
All throughout I maintained my fascination with natural history, the environment, and evolution. As I continued learning, I became more interested in mechanisms. My undergraduate degree was in evolutionary biology with a focus on physiology, and my graduate degree was in molecular genetics. My first postdoc was in synthetic biology, where we engineered systems and delved deeper into understanding how biology works. During this time I also realized that academic science wasn’t for me. It’s not that there was anything wrong with it; I just wasn’t suited to it, as I was a bit too much of a dilettante.
During a second postdoctoral fellowship, at Baylor College of Medicine starting in 2011, I began to collaborate with machine learning researchers. I was conducting basic biological research and realized that the mathematical tools that machine learning provided could help us understand biological mechanisms and predict how biology works. If you can predict how biology works and then prove you’re right, it means you’ve truly understood something.
Though I was doing academic research, by now I knew I wanted to find a way of pointing the science and innovations towards a commercial purpose. My collaborations with the data scientists was eye-opening — I totally drank the Kool-Aid. I loved what we were doing, generating data to feed their predictive models. This was 2011, 2012. These were the very early days of what we now think of as AI in the life sciences.
One of the European data scientists I was working with had just finished his PhD in artificial intelligence and started a company called Genialis. He was looking for a US-based co-founder who understood the life sciences space to complement his AI expertise. My role was to think, okay, what are some of the biomedical problems that we could potentially solve with this new field of mathematics? That is how I joined Genialis about nine years ago. The rest is history.
This is the only job I’ve had besides working as a laboratory researcher or a professional cook. I joined the company as the Chief Product Officer, where I helped build the original software product. Later, I transitioned to the CEO role, leading us to become a biomedical discovery business rather than a pure software company.
Can you share the most interesting story that happened to you since you began your career?
Here’s a story I enjoy because it made me feel a bit like Jason Bourne. One important thing to note is that I’m committed to supporting my wife’s career. I take advantage of the flexibility that comes with being an entrepreneur and running a remote-first company, or at least one that is very flexible in terms of where we work. I set myself up wherever my wife’s job takes the family.
We started in Houston, where she works for a large energy company, then moved to the Bay Area. We’re now back in Houston, but between the Bay Area and Houston, we spent two and a half years in Kazakhstan. During that time I was running the company remotely. I used Zoom and other tools, while frequently flying back and forth to our European and US offices from Atyrau, Kazakhstan, located where the Ural River meets the Caspian Sea. I also had to fundraise, which was challenging in 2022. The markets were tight following the biotech collapse at the end of 2021. We had promising growth in 2021, but 2022 and 2023 were difficult. So, there I was, flying around while fundraising in a challenging market.
There were lots of Zoom meetings with Silicon Valley at three in the morning. Finally, we got everything worked out. We had two great lead investors, and we were ready to sign the papers. I ended up flying from Kazakhstan to Boston for a 15-minute signing meeting that had to be done in person and on US soil to secure the investment. The very next morning, I flew back to Kazakhstan. That was a pretty crazy moment in international travel.
Earlier, before we secured the investment, my wife, a friend, and I were trekking in the mountains of southeastern Kazakhstan. We climbed a 14,000-foot peak in the ice and snow, then climbed down the mountain, cleaned up at the hotel, then I rushed to the airport. I flew to Istanbul then to Geneva, caught a train to Basel to give a conference talk and meet with potential investors. I felt like a Jason Bourne character going from the mountains in southeast Kazakhstan to Geneva for a meeting, then to Basel for a conference, and back again — all in just a few days.
Can you please give us your favorite “Life Lesson Quote”? Can you share how that was relevant to you in your life?
I have two life lesson quotes. The first is from Mike Tyson: “Everybody has a plan until they get punched in the face.” This quote became popular in my world when we had our first child. You think you have a plan but when that first baby comes, everything gets turned on its head. It’s proved true over and over again. In business, we plan a big product launch, but you never know what will happen until the rubber meets the road — until that product hits the market, or you get “punched in the face,” so to speak. This applies to many significant business decisions, whether it’s fundraising, sales, or a product launch. You can plan all you want, but you’d better be adaptable.
The second life lesson quote is simply, “Do unto others as you would have done unto you.” Treat people nicely. It’s a very small world and industry. Our networks constantly collide, so be kind. Life is too short to work with jerks, and we take this seriously when hiring and building our team. We over-index on recruiting for values, knowing that building a strong company culture and working with people you enjoy is crucial.
How would you define an “excellent healthcare provider”?
I’m really interested in the notion of healthcare versus sick care. In the United States, we primarily have sick care. Most of the industry waits until people become ill, diagnoses them, and tries to treat them either with medicine or surgery. An excellent healthcare provider intervenes early on the preventative side, focusing on helping people maintain their health rather than simply addressing issues once they arise.
What are your favorite books, podcasts, or resources that inspire you to be a better healthcare leader? Can you explain why you like them?
My favorite book, or the most influential book for me and what I do, is The Emperor of All Maladies by Siddhartha Mukherjee. This is his first big book in a series. I like his other books too — The Gene and The Song of the Cell — but The Emperor of All Maladies is such an accessible history of our battle with cancer. The horrors in the details of how early cancer surgeons went about treating the disease motivate me every day to get up and think about how we can use precision medicine. How we can try to target the illness on a person-by-person basis, rather than using broad strokes — whether those broad strokes are drugs that make you sicker or taking a knife to the patient in unnecessary ways.
The healthcare podcast I listen to most is Stat News: The Readout Loud. The reporters are super engaging and funny, and it’s a useful way to get a bird’s-eye view of what’s going on at the pharma level of drug development. I also enjoy the Harry Glorikian Podcast. Harry is a very thoughtful individual and has great guests to talk about innovations in data, data science and wearables. I’m going to plug my podcast, Genialis’ Talking Precision Medicine. As the name implies, we mostly talk about precision medicine but also AI, data science, and other topics with a super wide range of guests. That’s been lots of fun.
Are you working on any exciting new projects now? How do you think that will help people?
In April 2024, Genialis announced a new product called Genialis krasID, which formally launched this past September. This is a machine learning algorithm that predicts which patients will respond and benefit from KRAS inhibitors, one of the hottest and historically most important areas of cancer drug development. The project around krasID is to scale this AI model to work for all relevant tissue types (lung, colorectal, pancreas), and all different mutational targets around KRAS.
This is going to help a lot of people, because 25% of all cancers are KRAS-driven. Selecting patients by the mutation alone means that only about a third of the patients who get the medicine are going to benefit. Our model can do significantly better.
The bigger vision of Genialis is this notion of a “supermodel” of cancer. What do I mean by a supermodel? I mean a collection of individual machine learning models that have learned all of the various aspects of cancer. By collecting enough of these individual models, we cover all of the relevant biology in cancer. Our vision is to have a machine learning model that can predict very precisely the best possible therapy for all cancer patients, for any drug target, for every available drug. It will be instrumental because then precision medicine will scale to truly every cancer patient.
Ok, thank you for that. Let’s now jump to the main focus of our interview. What are the primary barriers that underrepresented populations face when seeking healthcare?
There are a great deal of socioeconomic barriers. I’ll talk about the one that I understand the best — the data gap. Genialis is doing something about underrepresented populations that are almost definitionally missing from the data. If there’s a population that’s underrepresented in terms of their healthcare access, the healthcare system will have collected less data on them. As we become more data-driven in our approaches to personalizing, tailoring, and delivering healthcare, those patients will continue to be underrepresented. It will exacerbate the problem. You get this death spiral: being missing in the data leads to a lack of access to new treatments or delivery modalities, which in turn results in even less representation.
Genialis’ goal is to focus on being extra careful, overly deliberate, and putting in the extra effort to build datasets that reflect the world. We aim to go the extra mile to find data from underrepresented populations so that we can incorporate that into the training sets for our AI models. We want our AI models to work for the world. We want our new therapies to work for the world. And to do that, we have to be very deliberate about gaps in the data.
How can healthcare providers build trust with patients from diverse backgrounds, especially in communities that have historically experienced medical neglect or discrimination?
That’s a really great question. My sense is that there are two things that are absolutely key here. One is empathy. We need to listen. We need to ask better questions, rather than prescribing out of the gate and assuming we know what’s going on or what’s best for people. We need to listen. Take people’s self-reported history seriously. Find out not only what’s ailing them but what their history has been. Give them an audience to describe the neglect or discrimination and acknowledge that you take it seriously. Starting from a point of empathy is absolutely essential.
The other thing is availability. Sometimes you have to meet people where they are instead of insisting that they always come to you. One way to potentially do it is by giving medical care providers back more time. Artificial intelligence can help if we can relieve some of the busy work burden from doctors, nurses, physician assistants, and other practitioners so that they have more time to spend face-to-face with patients, more time to listen and convey their empathy. That could go a long way to building trust.
What role does cultural competence play in improving patient outcomes, and how can medical professionals be better trained to meet the needs of underrepresented groups?
I’ll refer to my answer to question eight. I don’t know whether a medical provider needs to be able to blend into someone else’s culture or even fully understand it, but they need to be able to ask questions and be patient enough to listen to the answers.
I believe this active and empathetic listening is key. Acknowledging past neglect and past discrimination gives people a platform from which they can tell their side of the story without feeling judged. We need to establish trust from the outset. It could be impactful here.
Can you share any successful strategies or programs that have been implemented to reduce health disparities and improve outcomes for underserved communities?
I am not an expert on the provider side. However, on the technology/industry side, we’ve seen several exciting companies either adopt initiatives or form with the goal of representing underrepresented groups. There are companies with specific missions to bring certain modern biomedical technologies, techniques, and approaches to places like Sub-Saharan Africa.
Genialis has recently begun discussions with a potential collaborator in India that is doing an ambitious job of bringing precision medicine tools, which have become rather widespread in the West, to India. But not just to India; they also intend to move into parts of South America and other parts of South Asia. Being very mission-driven is something that lends itself well to startup culture. I can’t speak to what hospitals are doing, but a lot of startup companies and tech companies are saying, “Hey, let’s start with the underrepresented groups. Let’s solve their problems, and that will bring value to the whole healthcare system.”
How can technology and telemedicine be leveraged to reach underrepresented populations who may face geographic or financial barriers to traditional healthcare services?
I believe that telemedicine and certain technological solutions, including potentially digital therapeutics but definitely digital diagnostics, could be hugely beneficial for the United States. The United States is a very large country with disparate population centers and everyone else is in between. There are huge swaths of the country that don’t have healthcare providers of one type or another. This is acute in mental health. And mental health is one of those areas where sometimes just talking to someone on the phone or by video conference can be very impactful. I’m bullish on the ways the move towards digital medicine and telemedicine, especially in our post-COVID era, can help break down some barriers to access. For more information on this, I would encourage people to go to the Digital Medicine Society’s website and read some of their resources. That’s an organization that’s dedicated entirely to this problem. I’ll provide a link as well to our podcast interview with our friend Jen Goldsack, who’s the CEO of the Digital Medicine Society.
As a “healthcare insider”, if you had the power to make a change, can you share 5 changes that need to be made to improve patient outcomes for underrepresented populations? Please share a story or example for each.
I’m not a healthcare insider, so let’s talk about startup tech. This is the part of the world that I know.
1. Build diverse teams. We want to build companies that have a diversity of opinion, a diversity of experience, and a diversity of professional contributions. As we’re building our tech companies and the companies that are going to bring solutions to healthcare, we should aim to maximize the diversity on as many axes as possible, especially diversity of experience and perspective.
2. Listen and start from a point of empathy. Numerous startups and tech companies pitch that they’ve identified a huge problem, they know the solution, and you should buy their solution. The question is, could we do a better job of asking more questions and listening to people about what the problem is to solve?
3. Focus on solving problems for underrepresented populations. This will translate into large business opportunities for the rest of the world. There’s an example in drug discovery where it’s tough to develop a business model for drugs targeting rare diseases because, by definition, not that many people have them. The market for any single rare disease treatment might be small, but some companies, like Healx in the UK — my friend Tim Gilliam’s company — have repeatedly shown that you can leverage technology to scalably and economically discover drugs for rare diseases. This can have a significant impact for a number of patients while translating into a great business. We need to focus on solving problems for underrepresented populations without being deterred a priori that it’s not a big enough market opportunity.
4. We need to move to precision medicine. This is not specific to underrepresented populations; this is for everybody. It is ridiculous that drugs are prescribed without doing the proper degree of testing to understand which drugs will benefit which patients, in cancer in particular, but also lots of other disease areas: immune disease, neurodegenerative disease, et cetera. We’re developing extraordinarily accurate molecular biomarkers. While molecular testing costs a few hundred to a few thousand dollars per patient, it can save lives, add years to lives, and save hundreds of thousands, if not millions, in unnecessary treatment.
5. The idea of ensuring that underrepresented populations are represented in the data. I would point out some frightening trends, like those described in the book Invisible Women, which discusses how women are historically excluded from data and decision-making and thus are an underrepresented population, yet they make up half the world. Women are especially underrepresented in healthcare data. You better believe this is also true for minority and ethno-geographically diverse populations. Let’s work harder to ensure these patients are properly represented in the data.What specific steps can be taken to ensure that medical research and clinical trials are more inclusive of underrepresented groups, and why is this important for improving overall patient outcomes
What specific steps can be taken to ensure that medical research and clinical trials are more inclusive of underrepresented groups? Why is this important for overall patient outcomes?
I believe it’s heartening, at least in the US, that the National Institutes of Health already require
certain representation. Guidelines are met in medical research and clinical trials. However, we can do better in clinical trials to build trial populations that look more like the real world, and especially for later-stage trials. I think this is important because we see a significant drop-off between clinical trial outcomes and real-world outcomes for the same medicines.
Undoubtedly, a big part of this discrepancy is that, in clinical trials, we try to enroll the healthiest sick people we can. In other words, we want people who only have the malady we’re trying to treat and nothing else. You very rarely find that in the real world, where patient populations are full of comorbidities and have really diverse clinical trajectories. By adopting precision medicine, we can have more tailored trials in which we find patients who share relevant biological characteristics in common, even if they are more diverse and represent different ethnogeographies and histories.
You are a person of great influence. If you could inspire a movement that would bring the most amount of good to the most amount of people, what would that be? You never know what your idea can trigger. 🙂
A majority of my work focuses on people who are very sick and developing medicines where we’re trying to save them, perhaps by adding months or years to their lives. Ultimately, I want people to be healthy and stay healthy. So, my movement would be one that tackles the four things we know keep you healthy. Get more sleep, have less stress, eat better, and exercise more.
Those four tenets — sleep, less stress, exercise, and better nutrition — are obvious, but they’re not always attainable, especially in socioeconomically disadvantaged populations. I’m calling for more of a social movement. We need to allow people more time to rest, better wages, and the end of food deserts so people have access to better nutrition. We need to provide more discretionary time or more free time, which people can use for things like exercise and socialization to relieve stress. These are the kinds of things that would impact people’s lives. If I had to sum it up in one thing, it would be: can we move to a four-day work week? If we as a society moved entirely to a four-day work week, or even if wage and shift workers only had to work four days, people’s health overall would improve dramatically.
How can our readers further follow your work online?
I try to post actively on LinkedIn, so check me out there. Alternatively, my company’s website, www.genialis.com is a great place to see what’s happening. And do subscribe to Talking Precision Medicine, on whichever service you get your podcasts.
Thank you so much for these insights! This was very inspirational and we wish you continued success in your great work.
Rafael Rosengarten of Genialis On 5 Things We Must Do To Improve Patient Outcomes for… was originally published in Authority Magazine on Medium, where people are continuing the conversation by highlighting and responding to this story.