Exposomics is the comprehensive study of environmental exposures throughout an individual's life and their impact on health, capturing everything from diet and lifestyle to pollutants with high temporal resolution. Exposomics is a distinct dimension from genomics; understanding the bridge between them provides a fuller picture of health and disease. Originating from the world's leading exposome laboratory at Mount Sinai Health System, LinusBio is revolutionizing the understanding of environmental impacts on health through its advanced exposome sequencing technology. Most notably, the company has developed a novel biomarker for autism spectrum disorder (ASD), utilizing a simple hair sample to aid in detecting developmental risks early, promising new paths for intervention and therapy optimization. With the company’s diagnostic aid tool receiving Breakthrough Device Designation by the U.S. Food and Drug Administration (FDA) and a CLIA certification, LinusBio is setting new standards in medical diagnostics, monitoring, and treatment for central nervous system disorders and beyond.
David Alvaro (DA): Can you relate the genesis of both LinusBio and the exposome concept that is the foundation of the company?
Manish Arora (MA): I like to refer to this as our “origin story.” Even as a Ph.D. student many years ago, I was fascinated by how the world of medical science was evolving. My passion has always been understanding how our bodies respond to the environment. When the sequencing of the human genome was underway, the focus was on genetics and genomics, but I believed there was another dimension worth exploring, and I predicted that genomics alone wouldn’t provide all the answers, especially for conditions like CNS disorders.
That question — what lies beyond genomics? — became the foundation of my research at Mount Sinai and eventually led to the founding of LinusBio and our groundbreaking platform for studying the exposome. Exposomics refers to the investigation of all non-genomic factors and how we respond to them. Unlike our genes, which are mostly static from conception, our exposome is constantly changing. For instance, what I eat in the morning and how my body responds to it can be entirely different by night. The key difference — and the key challenge — is the time component.
We can now capture that time dimension from a single strand of human hair. This provides the same amount of data that previously required 500–1,000 blood samples. While a blood test offers just a snapshot, we’re building a molecular movie — all from a single, non-invasive hair sample that can be easily collected at home.
DA: How broad is the application of your platform in studying disease? And how have you determined where to begin?
MA: One of the primary challenges with a platform like this is deciding where to focus to achieve the greatest public health impact. There are two principal areas where our work can make a real difference.
The first is knowledge generation, in collaboration with academia, government, and the private sector. We’re collaborating with pharmaceutical companies on phase II trials for new drugs and nutraceuticals. We’re also working with the Japanese government on a project that’s not focused on a specific disease but aims to understand how to keep the overall population healthy. In addition, we have academic partnerships that explore different diseases or even the broader trajectories of human health — essentially, what happens over a person's life as seen through the time dimension of our platform, this "molecular movie."
The second area is direct clinical impact — reaching patients who haven’t benefited from genomic or other types of omics technologies. Two areas that stand out for their potential to be revolutionized by this technology are CNS disorders and gut or GI tract disorders.
Take autism spectrum disorder (ASD). We’ve spent hundreds of millions of dollars studying it, and it’s become a public health priority because its prevalence is rising dramatically — from 1 in 10,000 when I was born to 1 in 36 today. While some of that increase is due to better detection, there are real underlying issues for this rise, and they can’t be primarily genetic. Genes take centuries to evolve, so if something is changing this quickly, there must be other factors at play. Beyond ASD, mental health is a significant public health area that needs a different approach — consider this, even if were to do a full genetic scan of a newborn, I have only a very minimal ability to predict their future metal health — whether they’ll develop autism, attention deficit / hyperactivity disorder (ADHD), or schizophrenia is almost impossible to predict from all that genetic information alone.
DA: I have teenage twins with ASD, and I remember how slow and difficult the diagnostic process was when they were little. How does your StrandDxTM ASD platform transform that diagnosis, and what impacts can earlier and better diagnoses have on clinical outcomes?
MA: When we first started looking into ASD, what really surprised me was that we were diagnosing the condition based almost entirely on behavior — how a child appears to a clinician, often documented through a questionnaire. We wouldn’t accept that approach for most other conditions. Contrast this approach with how we manage type 2 diabetes — biology is always central to the diagnosis, monitoring, and treatment of type 2 diabetes, even though behavior and diet play roles in the disease. However, with autism, there was no biological component in the diagnosis, just behavior, and that had a huge negative impact.
Right now, the average age for an ASD diagnosis in our country is between four and five years old. But the brain is most malleable in terms of learning the skills that are critical for socialization, language, and managing repetitive behaviors in the first year of life, meaning that we are missing a crucial window of opportunity. Our goal was to fill that missing piece — detect autism earlier, identify who needs therapy, and tailor existing therapies accordingly. We submitted all our data to the FDA, and for the first time, they granted us breakthrough designation as a biomarker that can detect autism as early as one month of age.
DA: Can you expand a bit on exactly how those tests are conducted and what the data readouts look like?
MA: When the test becomes available next year, we will simply send out a hair collection kit to the child's home when we get an order from a doctor. The family collects a hair sample at home — just a simple clip with scissors, no need to pull the hair, which can be painful. A significant advantage of using hair is that the sample is stable at room temperature, with no special requirements. During the COVID-19 pandemic, when people couldn’t visit hospitals or clinics for blood tests, hair testing continued without any issues. Unlike saliva, which was highly infectious during the pandemic and couldn't be mailed, hair testing offers what I like to call a "last mile advantage."
Once the hair is collected and sent to our lab, it undergoes a series of automated processes: robots wash, clean, and straighten the hair before opening it up like a book. This might sound simple, but it's incredibly complex because hair varies so much in shape and size. The robots can handle everything from thick hair to a baby’s fine hair, securing it and opening it without generating heat or damaging any molecular signatures.
We obtain temporal signatures from the hair in a manner similar to counting growth rings in a tree. By mapping along these "growth rings," we can look back in time and see a detailed history. In a tree, if a particular section is irregular, for example, it might indicate a period of stress such as a drought. In a similar manner we can use growth rings in hair to obtain a timeseries of information.
Using laser technology to zap each growth ring, extract material, and analyze it using mass spectrometry, we can achieve temporal resolution down to about two hours. The resolution is so fine that it essentially creates a molecular movie of chemical signatures across timepoints with no missing frames. It's like a molecular Fitbit, but one that measures thousands of different factors.
The real challenge is analyzing this wealth of data, which is why we've invested heavily in data science, machine learning, and artificial intelligence to convert it into actionable patterns. For example, I could ask you, "What's the pattern of your cholesterol?" which might seem nonsensical, but to your physiology, it’s a crucial question. We can get an hourly pattern of your cholesterol over months, and if we notice a shift from a stable pattern to something more chaotic, it might indicate that it’s time to see your cardiologist. This is just one example of how our system works — from robotic processing to data science — ultimately linking it all to actionable insights in the clinician’s office.
DA: To what extent are you able to correlate those data with real outcomes in the conditions you’re studying?
MA: That's a critical point. We have to be very strategic in our approach, which means being laser-focused on our priorities. One of our top priorities is ASD, but more broadly, our focus is on brain health — particularly CNS disorders — because these are areas where there are currently no biomarkers, such as dementia or schizophrenia.
To support this focus, we've conducted numerous multinational studies, many funded by the NIH (National Institutes of Health). We’ve even done studies among identical twins in Sweden, where one developed autism and the other did not, despite sharing identical genetics, the same mother, the same socio-economic background, and sometimes even the same placenta (in identical twins). Clearly, there’s something beyond genetics at play, and it turns out it’s hidden in the time dimension. We’ve also conducted studies on newborns in Japan and older kids in New York, and we’re wrapping up a statewide study in California, with another just starting in Mexico. These are large, multinational studies, and what we're seeing is that we can identify who has autism and who doesn’t with a high degree of certainty. We’re even beginning to subtype it: for example, determining whether someone has both autism and ADHD, just ADHD, or just autism.
Many brain disorders are really a conglomeration of different conditions that get grouped under one umbrella because they manifest similar symptoms. Beneath those symptoms, there are distinct biological subclusters. Our ultimate vision for ASD, from both a public health perspective and a scientific one, is to have a meaningful impact by using our novel platform to redefine these disorders biologically.
DA: Another challenge with a disorder that is defined as a spectrum but may really be distinct subclusters is in finding the right drugs to treat an individual, which to date has been a process of trial and error to discover what is effective. Do you see exposome profiling helping redefine things from a pharmacological perspective?
MA: That’s the direction in which we’re heading. However, I want to avoid past mistakes, where we’ve often looked for a "magic bullet" — a single thing that will solve all problems –– which is never really the case. Biology is always more complex. I believe that our platform will provide a significant missing piece of the puzzle and help resolve some of that complexity. But it won’t eliminate all of it, which is why we always work in conjunction with clinicians. Our tool is designed to aid clinicians, not replace them — much like any other diagnostic test.
Our goal is to be a companion diagnostic tool, offering a new kind of precision phenotyping. Essentially, we want to be able to say, "This drug is a match for you," similar to what’s done in oncology — we sequence tumors to determine which chemotherapy cocktail is most suitable. The challenge with CNS drugs, as you mentioned, is that patients often go through a long and frustrating process of trying different drugs. Another issue is that the side effects of CNS drugs are immediate, while the benefits could take a month or more to appear, leading to poor compliance. Our temporal profiling aims to remove this messiness by providing a much more accurate picture of a patient’s physiology.
DA: Beyond ASD, can you expand on the rest of your pipeline and the road ahead for those programs?
MA: Our ASD program is quite advanced — we’re at the clinical stage. We’ll soon be rolling it out in some statewide testing programs and partnering with clinical and hospital networks in the United States, as well as Japan and the Middle East, where there’s a significant need. Over the next six months, you’ll start hearing more about these initiatives.
Next in our clinical pipeline are conditions that often co-occur with autism, such as ADHD. Many neurodiverse individuals have both autism and ADHD or one of the two, but they can be clinically confused because they present similar symptoms. We’re working to address these complexities.
As we continue along this trajectory, our focus will expand to other CNS disorders, which often emerge in late adolescence or early adulthood. Later in life, we’re particularly interested in ALS (amyotrophic lateral sclerosis). We have been communicating with the FDA about a biomarker for ALS and will be publishing our findings soon. This work stems from a U.S. study that we are now expanding internationally. As you know, developing a drug for ALS has been a challenging process — one drug was approved, only to be later unapproved, largely because there’s no reliable biomarker to measure efficacy.
Our goal is to support the development of new drugs, especially for late-stage diseases like ALS, dementia, Parkinson’s, and Alzheimer’s. We are actively seeking partnerships to advance these efforts. We’ve already seen success in our pharma partnerships, including two phase II trials — one with Novozymes and another with a large pharma company. Both have yielded promising results.
We’re also working with four universities — one in Europe and three in the United States — that are eager to access our platform. Additionally, we’re collaborating with the Japanese government on a project to profile the entire population. Our long-term vision is to map the human exposome on a global scale.
As we build this global data set, we’re prioritizing CNS and gut disorders because they lack biomarkers. For example, why do I have to get a colonoscopy rather than a simple blood or hair test? It’s because we haven’t developed the necessary biomarkers yet. So, while we’re creating this global data set that will eventually become invaluable, we’re doing so with a strong focus on prioritization and good corporate governance.
DA: Do you envision that tests for conditions like ASD and ADHD will eventually become standard for every newborn?
MA: That’s precisely the vision: to test every newborn for these complex disorders within the first six months of life, and to do it noninvasively from home. We already have heel-prick tests where we collect a small blood sample on a filter paper — the Guthrie card — which is used for genetic screening. That test covers 1–2% of prevalent disorders, and for those who test positive, it can be life-changing. You might learn, for example, that by taking certain steps, you can prevent the full-blown development of a disease.
When it comes to autism and ADHD, we’re already addressing conditions that affect 12–15% of the population, where we can make a massive impact right from birth — as you know, early intervention can have a profound benefit, and, while it requires time and resources, it has no major biological or physical side effects. Our long-term vision is exactly as you described: to test every newborn, chart their health trajectory from birth, and help them grow into the healthiest versions of themselves.
DA: I’m curious — you mentioned the Middle East and Japan — are prevalence rates relatively similar across the globe? Can this work help us understand regional differences bases on exposure?
MA: There are still a lot of unknowns. In Japan, the rates seem to be similar to those in Western Europe (~3%), where similar diagnostic tools are used for autism. However, in many other regions, we lack sufficient data because the diagnostic tools were primarily developed for English-speaking populations and then translated into a few other languages. These tools don’t fully account for cultural differences, which can significantly impact how social aspects of life are assessed — something that varies widely between cultures.
These tools aren’t a one-to-one translation; for example, if you simply run an English diagnostic tool through a translator, the Arabic version won’t necessarily work as well for an Arabic-speaking child as it does for an English-speaking child in the United States.
I want everyone to have access to a unified, standardized measure of where they stand. When it comes to molecules, there are no language barriers. All molecules essentially speak the same language, and at the end of the day, all life is built from just a few elements in the periodic table. That’s what I’m trying to do — unify testing across this common language of chemistry.
Similar to questionnaire tools, even electronic monitoring of behavior, such as eye tracking, must consider cultural nuances. In some cultures, children are told not to stare at their elders. In some regions, girls may wear veils, which cover their eyes. Their behavior and socialization are different because eye expressions aren’t as prevalent. In those regions, speech becomes more relevant. These cultural nuances aren’t captured by current technology, which is why it’s so important to bring biology into the diagnosis and treatment of neurodevelopmental disorders.
DA: As you strategize for the future, how are you balancing the potential to expand into more indications with focusing on the ones you’re already working on?
MA: That’s a challenge every company faces. We’ve always said we will be guided by the greatest public health need and where we can have the most significant impact. We decided early on that we would focus on making a comprehensive impact on ASD — from early detection and diagnosis to phenotyping, supporting clinicians, and even developing drugs and bringing them to market. When we tackle a problem, we aim to address it across multiple fronts. We’re not just a diagnostic company, and we’re not just an autism company — we are a platform company.
There are also times when opportunities arise unexpectedly. We might meet an excellent partner who brings something to the table that perfectly aligns with what we’re doing. In such cases, we might say, "Wow, this is something we can really help with." A good example is our work with Novozymes, where we recently completed a phase II trial. They’ve been a fantastic partner, particularly on a microbiome-related nutraceutical. We ran the entire trial, and it was a huge success.
DA: Is there anything else you’d like to share about the company, your financing, and what the next few years may bring?
MA: We recently closed our Series A3 funding round, and I'm pleased to report that we achieved a 30% increase in valuation, which is significant, especially in today's challenging capital-raising climate. Our team is expanding, and we expect to be a 50-person company by the end of the year. I'm fortunate to work with an exceptional group of individuals who share a deep commitment to our mission.
We are now truly an international company. We've opened offices in Japan and established a presence in the Middle East, reflecting our strategic growth and the global belief in our vision. Looking forward, we envision a world where every newborn is tested using our technology, facilitating a shift from traditional sick care to precision healthcare. The potential returns on investment are substantial — not just in financial terms, but also in terms of enhancing health outcomes. This technology is not a distant future concept; it’s becoming a reality.