“Now I’m in so much pain because I’ve got a totally screwed up back, degenerative osteoarthritis, degenerative being the operative word... God is punishing me for my past by keeping me alive with incredibly painful situations.”

That’s Ginger Baker, the drummer of Cream, describing his osteoarthritis. ~Half of American adults develop symptomatic knee osteoarthritis by the age of eighty-five, so the odds are that you will one day sit where Baker sat. When you do, your doctor might tell you to exercise and maybe lose some weight. Then, they might give you a gel to rub onto the painful joints, which could help somewhat. If it does not, then perhaps you will be given Aspirin or Tylenol, or steroids injected into the joint. And then, eventually, there is surgery to replace your knee.

All of this may make osteoarthritis a bit easier to live with, but often not enough. One attempt to improve this bleak situation was tanezumab, targeting nerve growth factor. At first glance, it seemed to work: It beat NSAIDs on pain and daily function. Then some patients developed rapidly worsening osteoarthritis, and the drug died.

Meanwhile, sprifermin (recombinant FGF18) took a different approach: stimulating cartilage growth. The drug produced about 0.05mm cartilage thickness versus placebo, but no improvement in how patients felt. Without an established threshold for meaningful MRI change in cartilage thickness, the structural benefit alone was not enough for approval.

We’ve had several partial successes, which are hard enough to come by. For complete success, it seems we have to combine multiple mechanisms in one drug. But how do we even start looking for that?

Lorecivivint, now submitted for NDA, attempts to do both, but its clinical effects have been inconsistent. It emerged from a discovery process built around Wnt modulation, with the pathway defining the candidate space from the outset. A search constrained to one pathway may find a promising drug, but offers little reason to believe it has found the best one. Better ones may simply have fallen outside the search space.

Our idea was to greatly widen the search space, apply the ideal drug profile as the search criterion, and then test candidates against each risk encountered in drug development. We applied that strategy by first putting hundreds of independently barcoded perturbations directly into diseased living tissue. After all our filtering, a candidate we code-named Omen 13 emerged and is now in IND-enabling studies. It was selected to suppress the inflammation associated with osteoarthritis pain, making movement more bearable, while also protecting cartilage from further destruction and supporting the production of cartilage matrix.

So, how did we do it?

Collecting omens

Drug development has been severely limited by the high cost of validating causation in high-fidelity models. This forces most programs to commit early to a small number of hypotheses (adding considerable risk of picking the wrong hypothesis), or to compromise by using less predictive model systems (which prevents proper derisking).

We avoid that trade-off by putting candidates (and combinations) into Mosaic Screening until we find the full therapeutic profile we’re looking for. But first, we had to choose the living system to run our screens in.

Most osteoarthritis (OA) research is done in young mice whose cartilage is surgically injured and degenerates within a few weeks. This system is fast, reproducible and experimentally convenient. But it models injured cartilage, not naturally occurring OA, which typically develops through years of loading and use. Unless the patient is an NFL player, OA rarely begins with one identifiable blow. Young mice also cannot capture the age-related changes that make us vulnerable to the disease. It would clearly be better if we could run our tests in disease similar to what patients experience.

So what would a more realistic model look like? Probably a large mammal with joints similar in size to ours, that lives long enough to develop OA naturally and progressively. It turns out that people who own horses, for racing or farming, see this happening all the time.

Shared disease etiology is necessary, but not sufficient. A useful model system needs to also resemble human OA at the cellular and anatomical levels. Naturally occurring OA in horses develops through years of joint loading, much like human OA and unlike surgically induced disease in mice. As a result, horses and humans show significant similarities on the cellular level, such as corresponding disease-associated populations of chondrocytes in the cartilage and fibroblasts in the synovium (we’ve presented this at a few conferences). Horses also have large joints with thick cartilage and an anatomy that permits much more direct sampling than a mouse joint.

The trade-off of using natural disease models with high predictive validity is experimental inconvenience, and greater heterogeneity across individuals (just like in human patients). Fortunately, Mosaic Screening cancels both of those: because it requires only a few horses, the increase in predictive validity quickly becomes worth this effort. And heterogeneity across individuals doesn’t matter when your control treatments are in the same animal as your candidates.

So we made a simple decision: use mosaic screening in this high-fidelity system until we found the right target. We no longer had to trade biological fidelity for screen scale.

Your destination might be closed when you arrive

The next question was what to put into the screen. Of course, biological evidence and human genetics can point us towards more likely targets, but we also had to consider whether a biological target could produce a drug.

Every time we enter a new indication, we first define the drug profile that would satisfy medical needs and market considerations. We can then choose to screen only targets that fit this profile. For OA, AAV gene therapy is an ideal modality for achieving local and lasting effects in the joint while avoiding systemic side effects. So we chose to screen therapeutic proto-assets directly: AAVs carrying constructs that already resembled clinical gene therapies, rather than targets that would need to be converted into drugs later.

Across a few rounds of screening, we tested an AAV gene therapy library containing 197 barcoded perturbations, including combinations. Each AAV contained a sequence designed to up- or down-regulate a candidate therapeutic target, along with a nucleotide barcode. We injected the library at a sub-saturating dose that kept the effects of individual perturbations distinguishable and the tissue environment unchanged (verified by sequencing non-transduced cells).

The result is a mosaic tissue, with a few AAV-transduced cells scattered among unperturbed, diseased tissue. The animal has effectively become a screening plate.

All together now

When we started Gordian, predicting physiology from single-cell transcriptomics was not a mature science (alas, it still isn’t). We had to develop ways to tie measured cell states to the physiological processes involved in disease. The story of how we build these therapeutic keys will be told elsewhere, but one important consideration was anchoring to human patient data to filter out animal-specific effects early.

We evaluated each perturbation using curated “molecular features”: groups of genes representing disease-relevant processes like cartilage degradation, inflammatory signaling, and oxidative stress (more detail in our recent preprint). For the first time, this gave us shared disease-relevant axes on which every perturbation could be ranked.

We ran the mosaic screen over one month, to ensure that we weren’t fooled by immediate effects but rather could compare the steady state patients would experience.

After four weeks, we extracted the cells, did single-cell RNA sequencing to resolve each one’s effects on the transcriptome, and nominated hits based on our physiology-informed transcriptomic features. Mapping the effects of every target in a high-fidelity model gave us a map on which every target could be compared against every alternative. This allowed us to compare the perturbed cells with control cells from the same animal and ask a relatively simple question: did the perturbation move its transcriptome towards a state that is associated with OA improvement? If it did, that is good. If it moved them in the opposite direction, that is also useful to know, although somewhat less exciting for everyone involved.

A simple way of looking at the data is as a matrix. Every column is a different perturbation, and every row is a different aspect of disease that is represented by a molecular signature. For each signature, we had already decided whether it should increase or decrease. This helped us rank perturbations by how strongly and consistently they moved multiple disease signatures in the desired direction:

Our causal validation pipeline narrowed the initial 197 targets to 32 candidates, which were next taken for confirmation in human tissue ex vivo.

Trust but verify?

Even though the screen was scored using molecular features of human OA, we still wanted to confirm that the effects were true across tissues from different patients.

Because we don’t have better therapeutic options, hundreds of thousands of US patients get knee replacement surgery each year. We got access to the tissues of the replaced knees, which let us test the candidates directly in patient tissue. We co-cultured cells and/or whole pieces of tissue, both the cartilage and the synovial joint lining, to preserve cross-talk relevant to disease.

Across eight donors, we collected lysates after five days (cells)/ three weeks (explants) and measured the amount of cartilage proteins (GAG) produced, as well as the protein levels of two major cartilage degrading enzymes: MMP12 and MMP13.

Of the 32 nominated targets, 13 showed sufficient cartilage protection to be promising disease-modifying candidates, toward making more extracellular matrix and breaking down less of what was already there. The top targets were then tested for anti-inflammatory effects on human synovium as well.

More than a feeling

OA involves cartilage degeneration, but it also causes pain. And frankly, patients care a lot about the pain. Because pain can’t be measured in human tissue explants, we expanded into an in vivo model designed to quantify it.

The standard pain model is called MIA, which uses a monoiodoacetate injection to injure the joint. Pain is then measured with a static weight-bearing assay which looks at whether the rodents shift weight away from the affected leg.

The positive control is a high-dose of two pain drugs: celecoxib and pregabalin. Both are used to manage OA pain in humans, making the combination a useful positive control in this experiment.

Omen 13 almost completely reduced the weight preference once expression begins, and the effect persisted through the three-week time point. As we mentioned earlier, the current standard of care for OA focuses on numbing the pain. This is desirable, but without disease modification, it only makes the path to joint failure and replacement less painful.

Heterogeneity is the spice of life

But pain relief was only half of what we wanted. We also needed to know whether Omen 13 could modify the disease itself, and whether that effect would persist across different biological contexts. We had already screened in a large mammal and validated in human tissue, so mice gave us a scalable way to test the effect of age and metabolism (relevant risk factors for OA), and build an efficacy profile for each candidate. We intentionally test in animals with progressive age-related OA and high inter-animal heterogeneity, rather than surgical OA induction, to better mimic patients.

We ran three studies in these mice: the first two used naturally aged animals, 19 or 24 months old, that had developed OA naturally. The third was done in obese mice, which also develop OA but faster, likely due to a mix of inflammatory and weight-bearing effects.

In all three, a single intra-articular injection of Omen 13 significantly reduced cartilage erosion and restored proteoglycan content, confirming that its mechanism remains effective and identifying patients with obesity as a potential expansion population.

Better than the Beatles?

Omen 13 looks very promising so far, but is it better than other drugs in the OA clinical pipeline? Clinical-stage comparators were included throughout screening and selection, including rhFGF18 (sprifermin) and Genascence’s AAV-IL-1Rα.

We had already seen that Omen 13 is differentiated from rhFGF18 in the screen where all 197 targets were benchmarked against it. But before advancing a candidate toward the clinic, we wanted to remove another layer of risk by comparing it directly with clinical-stage drugs.

Compared with both FGF18 and a new comparator, AAV-IL-1Ra, Omen 13 remained differentiated enough to encourage further development. It suppressed inflammation at least as well as AAV-IL-1Ra, while showing anabolic cartilage effects, and boosting SOX9, a central regulator of cartilage and chondrocyte formation.

Start Me Up

We went on a quest to find an OA drug that treats both degeneration and pain. We scrutinized hundreds of candidates in various human and rodent models on our way there, ex vivo and in vivo. Many showed promise pre-clinically, testament to the candidate abundance that we were able to create by scaling up causal validation with Mosaic Screening. But only one combined the particular profile we were looking for: anti-inflammatory, anti-catabolic, and supporting local therapy. That was Omen 13. It acted through multiple mechanisms, worked consistently across different human donors, and showed broader effects in head-to-head testing than a purely anti-inflammatory clinical-pipeline comparator.

We’ve had our INTERACT meeting with the FDA, which aligned us on a streamlined path to IND. Omen 13 is now in IND-enabling studies to support entry into clinical trials as a gene therapy…

Gordian Biotechnology

We’re building a machine that turns therapeutic target hypotheses into more time for people with age-related disease.

By Effie Klimi

Last year, Gordian launched Aristotle’s Apprenticeship with our first cohort—Diego, Juana, Arthi, and Gayathri (pictured left to right). Over the course of the year, they became integral members of our teams, contributing to the science while growing rapidly as biotech professionals.

We launched the program with a simple belief: the best way to learn biotech is to be fully embedded in it. That means doing real work, taking real ownership, and building the knowledge and habits that make you effective long‑term. The Apprenticeship is designed as a year‑long experience that combines hands‑on project work with structured learning, mentorship, and exposure to how a biotech company actually runs.

Subscribe now

As we prepare to welcome the next cohort, we asked our apprentices to reflect on the year: why they joined, what surprised them, and what they’d tell someone considering applying. A few themes kept coming up:

An On-Ramp for Curious Scientists

A lot of early-career roles assume you already have a well-defined plan. We built this program for a more common reality: you’re deeply curious, excited by ambitious science, committed to learning, and still figuring out what kind of biotech work fits you best.

For Arthi, the apprenticeship offered space to explore that question while directly working towards new medicines:

“I had graduated from college knowing I wanted to pursue a career in biotech, but I wasn’t sure the exact path I wanted to take.”

Juana saw the program as a bridge into industry:

“I had no previous experience in industry and only had research internships in academic labs. I wanted to pivot into biotech and thought the apprenticeship program would be a great segue.”

And for Gayathri, the year was a deliberate challenge before starting her PhD:

“I wanted a chance to dive into something new and challenge myself before starting school. I could tell the team cared deeply about their work and approached it with real rigor.”

Real Work, Early Ownership

The fastest way to learn isn’t to watch from the sidelines. From the start, apprentices join as full‑time team members and contribute to projects that matter. Ownership comes early and grows quickly.

Juana described how that looked in practice:

“In my first month, I was given a small project, and I was able to take full ownership while learning new assays. Over time, the scope got bigger. Each project was meaningful and impacted the company. I learned project management, presentations, and scientific writing.”

Diego echoed that same trajectory:

“In my first months, I was already developing data that was presented at Keystone [conference]. Over the course of the year, I took on individual projects and built out my own realm of ownership.”

Arthi reflected on stepping into an entirely new technical area:

“I helped contribute to the single-cell sequencing pipeline at Gordian and believe that the team’s great organization and mentorship allowed me to make important contributions early on. I had no prior experience, but gained an extensive understanding of sequencing principles.”

Gayathri immediately jumped in to build tools still in use today:

“I built computational tools that helped scientists at Gordian make decisions around target selection and analysis, making it easier to interpret across teams with different backgrounds.”

The “Wider Lens” of Industry Exposure

While project work anchors the apprenticeship, the program is intentionally broader than any one role. We designed it to be a holistic experience, one that builds not only technical skill, but also an understanding of the business of science.

Through Masterclasses, Journal Clubs, and the Brown Bag speaker series, apprentices get a front‑row seat to how biotech decisions are made and how science moves from idea to impact.

Diego noted that this structure prioritizes access:

“Program events prioritize face‑to‑face time with senior Gors, which really helps you integrate into the culture.”

For Arthi, those sessions provided valuable context beyond the bench:

“I’m glad I got to learn the other layers of the biotech industry that go beyond the lab and think about broader challenges companies face downstream.”

Juana’s experience was similarly expansive; she found herself in conversations with FDA scientists, authors, and company founders, gaining insight into a wide range of roles across biotech.

Subscribe now

Support Systems and Mentorship

You can’t grow quickly without support. At Gordian, asking questions, pulling someone into a problem, and learning out loud are all part of the culture. Each apprentice is paired with a manager and dedicated mentors, but support extends well beyond formal structures.

As Gayathri put it:

“It wasn’t just my immediate team; people across the company were generous with their time and advice. That made the year feel much bigger than just my day‑to‑day work.”

For Juana, mentorship also meant intentional career development:

“We had monthly ‘happiness 1:1s’ where my manager checked in on my goals, stress levels, and growth—not just the science.”

And Diego highlighted the value of everyday collaboration:

“Some of the most valuable moments came from borrowing someone from an adjacent team for a ‘two‑minute question’ that turned into insight I still use every day.”

The Culminating Capstone Project

All of this training and support leads to the Capstone Project, where apprentices are challenged to propose and execute a stretch project that improves Gordian’s science or strategic operations. These projects are far from theoretical; they are designed to have a meaningful impact on the company’s platform.

For Juana, the project stood out because “it felt like I was contributing more than just labor, I was contributing ideas among scientists” and allowed her to learn new skills and “become more familiar with our platform technology, project management skills, presentations and scientific writing.”

Meanwhile, Gayathri’s project focused on building something that “fits how people actually work.” She created a curation tool that introduced data-driven methods to represent biologically healthy and diseased signals.

By the end of the year, these presentations served as a true culmination of our apprentices’ growth, proving they were ready for the next step in their careers.

To the Next Class:

When asked what they’d tell future apprentices, the message was consistent: don’t self‑select out, and don’t be quiet:

Juana: Apply apply apply! Don’t discount yourself because you have little industry or academic lab experience.

Arthi: Ask as many questions and talk to as many people as possible.

Gayathri: Be thoughtful and thorough with everything you do… who uses it, how some results might be interpreted, and what context they need.

Diego: There is so much to learn from the people around you. Pay attention to how people go about their days, observe their habits, and don’t be afraid to try different things and find a cadence that works for you!

Interested in Joining the Next Cohort?

Gordian’s Apprenticeship Program is a year‑long, full‑time opportunity based onsite in South San Francisco, starting in June. It’s designed to pair real project ownership with mentorship, structured learning, and a front‑row seat to building science in an early-stage biotech.

And while the science is fast‑paced, we make sure the culture keeps up. We make time to build real relationships, whether that’s team bonding at the Exploratorium or at company retreats, because great science happens when people genuinely enjoy working together.

If what you read here sounds like the kind of year that would accelerate your growth, applications for this year are now open! And if you don’t check every box on day one, that’s okay. The right apprentices are the ones who show up with curiosity, initiative, and the willingness to become more capable than the day before.

Apply Now

Gordian Biotechnology

We’re building a machine that turns therapeutic target hypotheses into more time for people with age-related disease.

By Effie Klimi

Subscribe now

At Gordian Biotechnology, we believe great ideas can come from anywhere and that sparking new conversations is one of the fastest ways to move aging research forward. That’s why we launched our recent essay contest: to challenge scientists, entrepreneurs, and thinkers from all backgrounds to tackle one big question:

What could or should pharma research groups interested in aging biology focus on to accelerate the translation of research into therapeutics?

We received entries from across the aging biology community, each bringing a unique perspective and a creative take on how to turn promising science into real-world impact.

Among these, one essay stood out for its clarity, strong scientific grounding, and vision for near-term clinical translation: Soleil Ava Wizman’s proposal to target follicle-stimulating hormone (FSH) as a novel lever for healthy aging.

FSH levels spike dramatically during menopause in women and rise steadily in men with age. Research shows these hormonal shifts contribute to bone loss, metabolic dysfunction, visceral fat accumulation, and possibly neurodegeneration.

Soleil’s essay draws on genetics, epidemiology, and preclinical studies to make the case for FSH inhibition as an aging therapeutic. She highlights MS-Hu6, a novel antibody from the Zaidi Lab at Mount Sinai, which in animal models preserves bone mass, reduces fat, and improves cognitive performance.

Her roadmap: begin with postmenopausal osteoporosis, where the unmet need and regulatory precedent are clear, and expand into other age-related conditions that share this biology.

For her award, Soleil received a ticket to the 12th Aging Research and Drug Discovery (ARDD) Conference, where she’ll join a global community of scientists, clinicians, and innovators shaping the future of aging therapeutics. We’re excited to see her ideas spark new dialogue — and perhaps inspire the next wave of breakthroughs!

The full essay can be found Here.

Gordian Biotechnology

We’re building a machine that turns therapeutic target hypotheses into more time for people with age-related disease.

By Effie Klimi

This post was written by our CSO and originally published on our website.

A cure. Ostensibly the goal that scientists and biotech companies are working towards, but in our minds cures are often something mythical. Like fairytales, they happened once upon a time, and can’t be produced on demand like a car or satellite. The next cure can’t be predicted; it will come like a revelation, and we can only go through the rituals we believe the cure demands. Even investors in biotech companies often have an ingrained and indefinite pessimism, a sense that gambling is an inescapable part of the endeavor.

And yet we have cures. That’s why we live for many decades. We may not remember where these blessings came from, that it took sustained scientific effort rather than prayers or good fortune. But in aggregate, our progress is impressive:

Childhood leukemia went from <10% five-year survival in the 20th century to ~90% of children being fully cured.

Each of the millions of annual childbirths in the 19th century had a ~1% risk of death, a risk now reduced by ~99%.

Before the 1920s, Type 1 Diabetes meant a single-digit lifespan. Every single patient died rapidly. Now, a near-normal life can be had for roughly the price of a new phone per year.

Medications for heart disease and stroke have reduced mortality by two-thirds since 1950, with the frontline drug combinations costing about a single restaurant meal per year.

Subscribe now

None of these cures came about as a single stroke of genius. Each time we’ve reclaimed healthy years from disease, there was a societal decision that effort and investment were justified.

Cures are not fairytales or alien visitors. They are real, but not yet so abundant that we’re surprised when one is missing. When we go to the hospital, we still have to hope that a cure exists for whatever we’re suffering from.

Like waiting in traffic, our experience with disease is dominated by the absence of cures: when a cure exists we either avoid the disease or get treated quickly, and when it doesn’t we often live the rest of our lives with the disease. When it takes us decades to engineer effective treatments, very few people experience the needed cure arriving in their lifetime. This reinforces a sense that cures, like seasons, come in their own time.

It’s worth asking whether a felt lack of agency is holding back our rate of progress. Knowing that we are the source of every cure, could we ask for more? In competitions, from the Olympics to Kaggle to protein folding models, it’s common that when a record is first broken, many others soon achieve the same result (sometimes in very different ways, and/or without knowing how the record was first broken). When we believe something to be possible, we find a way to do it.

What would happen if society believed it possible to find a cure for any disease before it claimed a human life? How could we create a system that produces cures on demand, and why haven’t we done so already?

The dominant challenge is that biology is not logically structured, genes/proteins and functions do not map one-to-one, and there is no instruction manual. A body is a complex set of interacting feedback loops, and a ‘disease’ is not a cleanly defined process but emergent behavior from misaligned complexity. We can’t predict the outcome from a given prod well until we do it. And usually we can only measure the end state, which doesn’t tell us what happened to get there. Thus a cure means shifting a system we don’t fully understand to a specific desired state.

Biology’s most successful approach so far has been reductionism – taking the system apart to understand the pieces, then deciding which piece to prod. This works when we can find a single cause of disease, like a virus or cholesterol, but is extremely slow for complex, multifactorial diseases like dementia. But we can invert the process, by bringing our experiments into the complex system and using perturbations to understand it.

Intelligence will soon become incredibly cheap, making experimental data to interpret the scarcest resource. Biotechnology has enabled us to turn each individual gene up or down and measure the cellular response, even inside a diseased organ. Scaling this to perturb every gene, and combinations of genes, in each diseased organ is a question of engineering and optimization. With access to real outcomes, we can go beyond computational models making predictions based on imperfect understanding of disease biology. And sampling the responses to enough perturbations lets us map the physiological dynamics we don’t otherwise see, like mapping the ocean currents with thousands of floating GPS trackers. This gives us enough understanding to run the right test of possible cures.

Cellular responses can tell us which perturbations shift organs in the right direction, but to know how effective they are we need to measure how the human body responds. To do this at scale we can’t rely on the slow progression of symptoms – we need to be able to do a blood draw or body scan and know whether disease is retreating. Such biomarkers don’t yet exist, but we have the technology to make them: we can perform broad measurements of factors in our blood, both before and after disease. Proper mapping of such molecular changes to specific organs, and whether they drive or respond to disease states, would make a blood draw a window into the body’s response to a new perturbation. The same measurements would also tell us when people begin to develop disease, and let cures be administered in time.

The capacity to measure thousands of factors in a single blood sample means that the data from each sample can be used for multiple purposes. Cross-applicability lets our knowledge grow synergistically rather than linearly: each sample from a human test tells us how disease responds, and new understanding and predictive models can draw on samples from the past for confirmation. Each time we test to diagnose, we also create data for predictions. And vice versa.

Computational intelligence will transform the process of finding clues and connections in biology. Millions of research papers are published, far too many for a human to read. Absorbing this entirely, and connecting it to data from biobanks, clinical trials, and perturbational experiments will enable rapid hypothesis generation. Pooled perturbations in diseased tissues enables rapid testing of these hypotheses. And mapping this to high-dimensional measurements in humans, enables rapidly finding people who need the treatment and confirming their response to it.

Like the shift from scouring the Earth for organisms with medicinal properties to engineering proteins to purpose using biotechnology, the shift described above would be a new paradigm for finding cures. One that relies less on disciplines and specialized knowledge to understand each disease, and instead on an efficient and exhaustive search process that spans diseases. Achieving this is not guaranteed, not a matter of wishing or finding a magic trick – the components above are only briefly described but each involves immense challenges. Getting there requires us to truly believe that cures can be abundant, and dedicating ourselves to that effort.

Gordian Biotechnology

We’re building a machine that turns therapeutic target hypotheses into more time for people with age-related disease.

By Effie Klimi

This post was written by our CEO Fransisco LePort and CSO Martin Borch Jensen, originally published on our website.

If you’re reading this, you’ve undoubtedly noticed that AI has delivered some incredible advances in the last couple of years. ChatGPT can answer most questions with ease, programmers regularly use tools like Copilot to help them finish code or create it from scratch, fully self-driving cars are operating in several major cities, and anyone can create incredible digital art by typing a few words.

These amazing algorithms have been thrown at basically every industry, including biotechnology. Unfortunately, while success has been impressive in biotechnology outside of drug discovery (AlphaFold, Nobel Prizes, etc), success in telling us how to cure disease has been far from meteoric. Exscientia wound down its Phase I/II study, BenevolentAI’s dermatitis drug failed Phase IIa, and most recently, InSilico and Recursion reported Phase IIa successes but had little to say on efficacy. Recursion lost 16% of its value that day, and InSilico has since scrapped plans for an IPO, again. The headlines paint an uncertain picture at best:

AI Does Not Make It Easy

AI tools don’t equal guaranteed drug success as BenevolentAI flops dermatitis test

After years of hype, the first AI-designed drugs fall short in the clinic

With few efficacy details, Insilico claims a Phase 2a win in latest AI readout

Beyond the Hype: Unveiling the Real Impact of Generative AI in Drug Discovery

What gives?

This quote from the last article gives us a clue:

“While generative AI shows promise, the process of transforming an AI-generated idea into a viable therapeutic solution is a challenging task. AI can predict potential drug candidates but validating those candidates through preclinical and clinical trials is where the real challenge begins.”

This points to a fundamental, structural problem: Drug discovery has, and still does, take place in one system, but validates those discoveries in a different system.

Subscribe now

Think about it this way. ChatGPT’s output is language. Naturally, it was trained on the compendium of written language. Copilot produces code for programmers and was unsurprisingly trained on code written by programmers. Self-driving cars drive in traffic… and are trained on videos, radar, and lidar gathered from driving in traffic.

Whether a drug cures disease must be validated in a living organism (in vivo). But the need for efficiency has forced drug discovery to seek out simplified systems to run tests at scale1: cultured cells in wells, organoids, explants, tissue slices. AI and robotics have made this process much more efficient, and new ways to measure and perturb systems have been developed, but it is still the same fundamental discovery process. In AI speak, drug discovery models are being trained in the wrong system.

Your training data should look like your test data. You wouldn’t train a self-driving car on the California Vehicle Code. If you’re going to test your drug in a living mammal, you must train for discovery in that living mammal.

Redesigning preclinical R&D

Preclinical discovery is complex. Countless articles have described the process, so I will skip the details here, but this image pulled from a quick Google search summarizes the process in 12 (not so) easy steps. 11 of them take place outside of the preclinical animal model system:

A tell-tale sign that we haven’t taken maximal advantage of the incredible AI, robotics, and omics tools invented recently is that the companies leading this effort have preclinical processes that look essentially the same, just with AI and robotics added. Here is Recursion’s 17 step preclinical process. The in vivo validation animal model comes in on step 17, and reminds us that “validating those candidates through preclinical and clinical trials is where the real challenge begins.”

InSilico has 12 steps they’ve transformed with AI. Validation presumably comes before Medicine42:

Using AI to make existing processes more efficient is certainly an improvement, but leaves an enormous amount on the table.

A feedback loop built for AI

At Gordian, we’ve redesigned the preclinical process from the ground up. Since all drug developers must validate in animals, Gordian trains for discovery in animals on step one. Our preclinical discovery process takes place entirely in vivo (see The In Vivo Screening Revolution and Introducing High Throughput In Vivo Screening for more detail):

Subscribe now

Most importantly, our training data (in vivo screens) comes from the same system as our predict / test data (in vivo preclinical validation), allowing a feedback loop built for AI.

By training and screening in vivo, our predictions for in vivo results are automatically, dramatically better. This isn’t because we use better AI or more advanced assays. This is because training data comes from the same system as our validation system, cutting away the caveats and unknowns of applying conclusions drawn from one system to another.

We can make this training data even more predictive by screening in models that share traits like anatomy and origin of disease. The expense of in vivo experiments lies in the animal model, especially when going beyond young mice. Since we can screen thousands of therapeutics with high confidence, we can bear the cost of much more advanced animal models (we call these Patient Avatars). This directly improves clinical translation:

Finally, our platform can produce biologically validated therapeutics at scale. This means that rather than pushing forward a single target/drug that takes millions in sunk cost before assessing whether it looks good enough to advance to trials, we produce dozens of clinically viable therapeutics, rigorously evaluate their potential, and nominate the best for clinical trials.

Curing complex, age-related disease

Current applications of AI to drug discovery miss the forest for the trees. AI is used to make the existing drug discovery paradigm more efficient, but the opportunity to correct structural flaws in the legacy discovery process is missed. We believe this has led to the questionable success AI has had in improving drug discovery thus far.

If taken advantage of maximally, existing AI technology has the capacity to radically alter our ability to discover breakthrough drugs, particularly in the complex, age-related disorders that have stymied discovery efforts thus far.

Gordian Biotechnology

We’re building a machine that turns therapeutic target hypotheses into more time for people with age-related disease.

By Effie Klimi

This piece was written by Isabel Matteo, who interned with us at Gordian.

Traditional drug discovery and development has a history of being long, painful, and expensive. Companies like Gordian are extracting knowledge from this decades-long history, and are using it to create something much better in their mission of curing age-related disease.

This summer, I interned at Gordian to understand the R&D philosophies and best practices at biotech and pharmaceutical companies, in order to reveal practices that may not be intuitive, but where their understanding may improve Gordian’s own R&D. Basically, what are the most productive R&D practices?

Subscribe now

But how is R&D productivity actually measured? The formula below correlates R&D productivity with five different levers, and is used broadly among large biopharmas to measure productivity.

Abbreviations: Productivity (P), work in process (WIP), project success rate (PTS), portfolio value (V), cycle time (CT), cost (C)

As for how specific companies optimize for productivity, we knew of one paper articulating AstraZeneca’s R&D philosophy, but wanted to see if any other companies were laying out similar research philosophies, and where they disagreed. After perusing papers, websites, and slideshows, I was able to find three core R&D frameworks: Eli Lilly’s “Chorus model,” Pfizer’s “3-pillar framework,” and AstraZeneca’s “5R framework.”

These papers outlined step changes in their R&D processes and highlighted the positive effects that they had, but I still wanted to know: how do these biopharmas actually implement their strategies, and how much of what they preach is actually useful in influencing R&D productivity? To answer this question, we can look at each of these company’s R&D philosophy changes in accordance with the annual numbers of drug approvals and phase 1 clinical trials.

An ideally productive biopharma would have the least number of clinical trials for the most approved drugs yearly, and it should also be noted that the average time a drug spends in the clinic before it is approved is around 7 years. Still, there does seem to be a clear correlation between the occurrence of R&D philosophy shifts and the difference between numbers of trials and drugs in development for these three pharma companies. AstraZeneca and Pfizer seem to have tightened their requirements for starting clinical trials, leading to fewer trials but with a higher average success rate (approvals unchanged or increased), while Eli Lilly seems to favor initiating more phase 1 trials, presumably with an emphasis on killing ineffective programs early.

Subscribe now

In addition to analyzing publicly available data, I reached out to 75 people in the biopharma industry, and asked for their insights on how R&D philosophy shifts actually happen at these companies, and what the most impactful decisions have been. From the 9 amazing people who agreed to interviews, I was able to piece together these 4 best practices for “productive” R&D in biopharma.

1. Establishing PK/PD markers early and at multiple stages in the discovery process, in order to understand the biology risk before advancing to later stages.

2. Getting to the “killer experiment” in the first ~5 months of discovery, with an objective process for determining the experiment successes and failures wherever possible.

3. Reducing cycle time and cost by trimming therapeutic areas to <5 and incorporating a ranking system so that highest priority projects are completed first.

4. Understanding the market for a product early in the discovery process, and incorporating studies to test the differentiation of a product against the standard-of-care treatments.

This is by no means a foolproof guide for biopharma R&D operations, and biopharmas often have conflicting views on how a budget should be used or the best organizational strategy, among other issues. And every company has specific considerations based on their indications and pipeline. But Gordian will integrate these lessons into R&D strategy, alongside our unique strategies to improve clinical translation.

On a more personal note, before starting this internship, I was on the phone with Martin when he asked me if cold-emailing people was something I’d be comfortable with. I, like most of my generation, don’t enjoy sending emails to strangers, especially those with more power than me. Then again, every period of growth in my life has come from doing something that’s scared me. And so, I answered: “I’m not comfortable with it now, but I guess I’ll have to be.”

That decision paid off.

Over the past 3 months, I’ve gone from knowing little-to-nothing about the biopharma industry, to being able to understand the R&D process and how each company navigates it differently. In this time, I’ve learned that at worst, an email goes unanswered, but at best, I’ll get to learn from an expert on topics we’re both passionate about, and develop a lasting connection. Most importantly, I’ve learned what it’s like to work at a company run by people who love what they do, know where they want to go, and will work incredibly hard to get there.

In my (albeit short) professional career, I’ve never before seen a company with the kind of culture Gordian has. Gordian fosters curiosity—with an office library filled by books on R&D, gene therapy, and more. Excellence is an expectation for all the work that is done here, but it never affects the camaraderie among fellow Gors. Overwhelmingly, the people here possess unmatched levels of kindness, whether during lunches—where friends checked in to see how my marathon training was going—or in cheering me on through my final presentation. I’m so grateful for my experience here, and I sincerely thank everyone who has helped me through it.

Gordian Biotechnology

We’re building a machine that turns therapeutic target hypotheses into more time for people with age-related disease.

By Effie Klimi

A handful of innovators have developed new, high-throughput in vivo technology that has led the industry to the threshold of a radically different world of drug discovery. This technology will have a game-changing impact on tackling complex and age-related diseases.

Life is a complicated matter. The number of specialized components working in concert to keep an individual cell alive is mind-boggling, let alone the interconnected functioning of the trillions of cells that make up a human. Failures of biology can likewise be complicated. We’ve made great progress in finding solutions to diseases with simple dynamics, such as bacterial infections, largely by scaling up and automating cell culture systems. But this approach has struggled to produce cures for diseases involving more complex dynamics, likely due to the limitations of these ex vivo, cell culture models.

The most conspicuous absence from ex vivo models is a functioning immune system, which is too complex to model ex vivo. For instance, the immune response to cancer cells involves not only the direct attack by cytotoxic T cells but also modulation by regulatory T cells and the influence of cytokines from different cellular sources. This means that a wide range of diseases modeled ex vivo are missing either their natural defense, like clearance of pathogens or tumors, or their primary driver, as in autoimmune destruction of tissue in diabetes or sclerosis.

Subscribe now

Immune infiltration and inflammation are also major factors in chronic fibrotic disease.

In e.g. pulmonary and hepatic fibrosis, disease progression is driven by bidirectional feedback loops between fibroblasts and immune cells. Activated macrophages secrete TGF-β and other factors which stimulate fibroblasts to differentiate into myofibroblasts. These myofibroblasts produce extracellular matrix components, leading to tissue scarring. In turn, the extracellular matrix and other factors secreted by myofibroblasts further activate macrophages, perpetuating the cycle of inflammation and fibrosis. This reciprocal interaction is crucial for the development and progression of fibrosis and is difficult to replicate ex vivo, where the dynamic and spatial interactions between these cell types are absent.

Signals between different cell types are important in every organ, and physical coordination is common as well: synapses between neurons, barriers formed by endo- and epithelium, and heartbeats all rely on physical connections between cells. Electrical, biochemical, and physical signals at the interfaces of cells serve to coordinate cellular actions. Failure of coordination leads to conditions such as atrial fibrillation. Ex vivo models are being developed to replicate such mechanical stressors, but so far, we can only replicate a small local section of an organ.

These are just examples that we know about. New mechanisms and interactions are discovered every year, so it is obvious that we are nowhere near fully understanding physiology. Thus, even the most carefully constructed ex vivo model system risks leaving out aspects of biology we didn’t realize were important.

The reality remains that the only model systems that allow for the inclusion of all relevant known and unknown biology for the disease in question are living diseased animals.

Why does this matter for drug discovery?

Drug discovery aims to identify therapeutics that work in living humans. For obvious reasons, confirming that newly discovered therapeutics are indeed efficacious and safe in humans must be done with the utmost of ethical consideration. This is the reason behind carefully regulated clinical trials. Prior to trials, evidence is built to motivate the expense and risk associated with clinical trials. This preclinical work is split into two central foci:

Advances in high-throughput screening have enabled rapid and cheap exploration of therapeutics in human cell lines. Because these cells are easily manipulated outside a living animal, millions of experiments can be run in parallel. But these experiments lack the in vivo environment described above. Thus their results must be validated in a living animal. So far, this validation process has been slow and expensive.

Because of this relationship, ex vivo testing has traditionally taken place first, narrowing the funnel of potential therapeutics going into in vivo testing in order to minimize total cost. The results of in vivo testing are fed back into ex vivo testing, and the loop progresses until a suitable therapeutic is discovered. Drug discovery thus typically progresses along three stages:

1. High-throughput screening: Millions of therapies are tested in parallel ex vivo, in order to identify molecules that engage the target.

2. Hit to lead: Promising therapies undergo limited optimization ex vivo to enable suitability for in vivo testing.

3. Lead optimization: Therapies are further optimized in order to maximize in vivo efficacy. Results of ex vivo optimization experiments are fed into in vivo validation experiments, in order to iteratively arrive at an optimized therapy.

Once discovery is complete, the optimized therapeutic undergoes IND (investigational new drug)-enabling studies and enters clinical trials. The progression of discovery and associated costs per phase are described in S. Paul et al, Nature Reviews, Drug Discovery, vol 9, March 2010 and summarized below:

Of note, the loop between cheap ex vivo testing and costly in vivo testing results in significant expense, particularly in the lead optimization phase.

The power of in vivo screening

High throughput in vivo screening ushers in a new paradigm. In this case, millions of experiments can be run in parallel, in the full context of the in vivo biology necessary for ultimately vetting therapeutics for clinical trials. Further, every therapeutic hit is, by definition, functional in vivo and thus requires relatively little optimization post-discovery. There is no ex vivo / in vivo loop, rather everything is complete in vivo in one step. The best of both worlds is achieved.

Several companies and institutions are working on enabling high throughput in vivo screening. Gordian’s approach (described in our introductory blog post and in more detail in our white paper) identifies novel therapeutics for complex age-related diseases, and results in over 3x improvement in time required and 10x improvement in cost required per IND-enabling study

Other companies are also researching pooled in vivo screening for different problems in drug development. Manifold Bio and Waypoint Bio are working on delivering cancer therapies to tumors with high specificity, for protein drugs and cell therapies respectively. Vevo Therapeutics is testing cancer drugs against pools of cells from different patients implanted into living mice, so that test results include mechanisms of drug resistance that span human heterogeneity, and depend on interactions with the cellular environment.

Subscribe now

Faster, cheaper, and better: the impact of in vivo screening on clinical development

The cheaper cost and faster timeline of discovery enabled by in vivo screening have a massive impact on drug development as a whole. Because the majority of the cost associated with drug development is in running clinical trials, maximizing the probability of clinical trial success is imperative. High-throughput in vivo screening enables preclinical exploration of orders of magnitude more targets and therapeutics for a given condition. The bar for moving out of preclinical research and into IND-enabling studies and clinical trials can be made much higher, allowing only far superior therapeutics to advance.

Even further, in vivo pooled screening enables the use of more advanced animal models that offer better clinical translation. Most traditional in vivo studies are done in models where disease-like pathology is induced rapidly and artificially: tumors are implanted into non-cancerous tissue, toxins are used to simulate degenerative disease, organs are damaged or removed using surgery. While these accelerated models indeed provide an in vivo test environment, this environment is somewhat artificial, and thus carries many of the deficiencies of ex vivo tests.

On the other hand, using in vivo screening, only a few animals are required to produce data that otherwise would have required thousands of animals. This loosens logistical and financial constraints, enabling the use of (costly) large or advanced animal models that have developed disease progressively and/or spontaneously, much like the patients they are intended to model. For example, Gordian’s Osteoarthritis in vivo screens utilize naturally occurring horse models, rather than the mechanically induced mouse models typically used.

This means in vivo screening enables experiments that were once practically impossible. Screening 100 novel targets in naturally occurring horse models of osteoarthritis would require approximately 1,000 horses using traditional methods (10 horses per intervention cohort). Even disregarding the extremely large expense of such a study (several tens of millions of dollars), there is no institution on the planet that could support such an experiment. On the other hand, generating analogous in vivo data from a high throughput in vivo screen requires two horses, and costs approximately $500K, an experiment Gordian has already performed several times over.

The revolution is here

At Gordian, we see high-throughput in vivo screening as the transformative evolution of traditional (ex vivo) high-throughput screening. The technology unlocks enormous possibilities across all areas of disease biology, which will be critical to discovering the next set of cures for complex disease. If you’re excited about this potential, please consider joining or partnering with us!

Gordian Biotechnology

We’re building a machine that turns therapeutic target hypotheses into more time for people with age-related disease.

By Effie Klimi

This year’s ATS meeting was inspiring in terms of progress in research and treatment, and also in how much more is needed. Our biggest takeaway is that idiopathic pulmonary fibrosis (IPF) is a tough nut to crack, and patients need better treatment options. For better or worse, the IPF community across the board recognizes that the existing therapies are simply not good enough for patients from both a survival and quality of life perspective. In addition, nucleotide/AAV gene therapy for lung disease continues to lack significant traction.

The meeting reinforced the core belief on which Gordian’s lung program is founded:

1. Patient matching novel preclinical models are needed that closely resemble human disease features, both pathological and -omic.

2. Testing the scores of potential targets that patient predictive models produce in such models will be required to maximize probability of translation.

Gordian aims to bridge this gap by identifying human genetic targets that are active drivers of the disease in preclinical models. These targets can be coupled with existing delivery technology to produce novel therapies for prevalent, age-related disease.

Subscribe now

Here are our highlights from #ATS2024:

Progress developing therapies for fibrotic lung diseases

• We were thrilled to see positive results presented from three promising phase 2 clinical trials (Pliant, Endeavor, Vicore). Notably, two of these trials were presented by Dr. Toby Maher, an advisor to the Gordian lung program.

• Insightful discussions from clinical experts shed light on the data from phase 3 IPF trials and the valuable lessons learned from failures. We learned that phase 2 trials often lack representation from general patient populations (ZEPHYRUS-1 and ISABELA 1 and 2).

• Academic and industry discussion around the need for better patient segmentation based on omics-based technology. Duke University’s efforts in identifying proteomic markers using samples from the Idiopathic Pulmonary Fibrosis–Prospective Outcomes (IPF-PRO) registry were particularly encouraging.

• There was an emphasis on the need for composite endpoints to assess treatment efficacy in IPF clinical trials by focusing on how the patient “feels, functions and survives.” There was significant support for designing future IPF clinical trials to include endpoints mirroring the lived experiences of IPF patients by including various clinical outcomes (e.g., respiratory hospitalization and patient-reported outcomes) rather than relying solely on forced vital capacity (FVC)to determine the outcomes of IPF clinical trials.

What’s new in preclinical IPF research

Traditionally, the IPF field has relied on single-dose bleomycin models and simple 2D fibroblast cultures for drug discovery. However, at ATS2024, we saw a big push for more sophisticated models to improve the translation of preclinical results into clinical success.

• Precision-cut lung slices (PCLS) from both diseased and healthy human patients are emerging as a key translational model for IPF. Many academic and industry experts are now focusing on using PCLS as a go-to model for testing drugs.

• Significant advances in fundamental methods related to the PCLS model were discussed, and we have already begun implementing these learnings by processing human and rat lung slices using current best practices.

• Recognizing the shortcomings of the single-dose bleo model, our industry colleagues are developing variations of the chronic-dose bleo model and aged models and validating these models with standard-of-care drugs.

• Single nucleus RNAseq of PCLS is increasingly used as an endpoint to study the mechanism of action (MOA) of therapies in human systems.

Gordian Biotechnology

We’re building a machine that turns therapeutic target hypotheses into more time for people with age-related disease.

By Effie Klimi

Five Gors descended on Baltimore last week for the 24th American Society for Cell & Gene Therapy Annual Meeting (#ASGCT24). As usual, we came back full of knowledge and inspiration, and thoroughly exhausted.

With thousands of attendees and many speakers and companies, there’s a massive amount of information to process. Following are highlights through the Gordian lens:

Subscribe now

Progress Developing Therapies for Prevalent Disease

A number of companies shared promising trial results, including:

• Sardocor’s phase 1 HFrEF clinical trial using AAV1

• AskBio’s phase 1 clinical trial using AAV2i8 for congestive heart failure now moving to a phase 2 trial

• Pacira Biosciences’ AAV2.5 ILR-1a trial in osteoarthritis

• Korean gene therapy company @ICM Co Ltd’s phase 1 AAV5.2 Nkx3.2 osteoarthritis trial

Trends in AAV Innovation

• Understanding AAV genomic integrations as both something bad due to off-target effects or potentially good for sustained expression

• Understanding and engineering inverted terminal repeats (ITRs) – using natural AAVs and not just AAV2 ITRs, looking at the effects of deletion and proximal regions

• Engineering capsids and promoters to de-target the liver for more specific gene delivery

• Thinking outside of the box for improvements to AAVs
  • Cirsium Biosciences manufacturing AAVs in plants (!?)
  • Regeneron using antibody-conjugated AAVs as a platform

New Knowledge in AAV Manufacturing

• New innovations in the AAV polishing step of removing empty capsids, with many people using the Sartorius CIM Monolith columns

• More people are talking about partial AAVs (vectors with truncated genomes)! Spark Therapeutics showed purification of empty, partials, and fulls through oversaturated loading multicolumn AEX. People were excited about the Refeyn SamuxMP mass photometer as the instrument for AAV assays

• Gordian has noticed vectors with truncated genomes as well, observing variable truncation rates depending on genome length, sequence, and structure. We found that charge detection mass spectrometry (CDMS) shows that prolonged exposure to acidic conditions causes viral genome extrusion

• Lukas C. Bachmann and his team at the Salk Institute optimized a method to successfully package and express transcripts for large proteins in AAV through RNA-end joining. This is an important step forward for reliable expression of large proteins in gene therapy.

• Tomoko Hasegawa and the team at UNC-Chapel Hill tested the efficacy of their novel immunomodulator based on HLA-G in a high-risk corneal transplant model. The results encourage further study in less immune-privileged tissues. Against control, the treated animals suffered no transplant rejection.

Advances in CRISPR Targeting

• Profluent trained LLMs on 500m proteins and 240k type II Cas effectors to generate novel proteins with comparable on-target activity and improved specificity to SpCas9 (this can be applied to gRNAs and other proteins too)

• Epitor Therapeutics unveiled CasNano, the smallest known CasEffector date, at just 399 AA

• Epic Bio presented posters on small yet potent cell type specific activators that can outperform VPR

• Many talks and posters focused on prime, RNA, and base-editing as approaches to minimize lasting off-target effects in the clinic

Thanks to the organizers, speakers, posters, and all attendees for an incredible meeting. We look forward to ASGCT2025 in New Orleans!

Gordian Biotechnology

We’re building a machine that turns therapeutic target hypotheses into more time for people with age-related disease.

By Effie Klimi

This post was written by our CEO Fransisco LePort, originally published on our website.

Creating Time

You want to wake up each morning, more capable than you were the day before. But regardless of genetics, and physical or mental self-care, a human existence has inevitably become constrained by time. Days become filled with hospital visits, and simple tasks like walking up stairs or hugging your grandkids become a struggle. Gordian Biotechnology envisions a life limited only by our imagination rather than by our physiology, one where we continue growing wiser and stronger, one that gets better each day, both mentally and physically. We produce medicines that allow each of us to look forward to getting old. We are creating what all of us are running out of: time.

Subscribe now

Drug discovery is hard, especially for age-related disease

Animal testing is currently the bottleneck in drug discovery. Until now, animal testing meant we could only make one guess at a curative therapy at a time, and testing those guesses was very expensive. Moreover, age-related disease is incredibly complex, so our guesses have been incredibly poor. Together, this has led to the stream of clinical failures we have experienced for nearly all age-related disorders.

To solve this, we must be able to make thousands of guesses, each guess needs to be cheap, and we must test them rapidly.

Mosaic Screening

The concept is simple: rather than experiment with perturbations to cells in a dish then move best guesses into animals one at a time, perturb cells directly in a living animal. A dish makes things easy: cells are accessible and plentiful, different perturbations are separated by walls of plastic, and the system is compact and highly reproducible. But a dish doesn’t replicate the living system. It lacks an immune system, blood, hormones, and the full suite of cells in the 3D structure of living tissue…a dish lacks nearly all of the things that make up an aged environment.

At Gordian Biotechnology, we experiment on cells directly inside a living, aged animal with the disease of interest. A pooled library of gene therapies is introduced into the target tissue, turning it into a “mosaic.” Various cells receive different individual therapies, each with a unique barcode, separated not by walls of plastic but by diseased tissue unperturbed by any treatment.

We single-cell sequence the treated tissue and use the library’s unique barcodes to identify the specific treatment given to each cell. The single-cell transcriptome allows us to explore various aspects of disease biology simultaneously in an environment very similar to where a clinical drug would need to be effective.

In this way, we run the most patient-predictive in vivo experiments, at a scale previously thought impossible.

Link to video

80% Predictive Power

Mosaic screening is operational across three diseases: osteoarthritis (OA), metabolic dysfunction-associated steatohepatitis (MASH), and heart failure with preserved ejection fraction (HFpEF), with applications to pulmonary fibrosis under development. The list of areas we can tackle next is long: neurodegenerative, neuromuscular, ophthalmologic, renal, intestinal, etc.

Our screens take advantage of advanced animal models we call Patient Avatars, and have each explored up- and down-regulation of hundreds of gene targets and target combinations:

Our OA screens are done in horses with naturally occurring diseases. Using our AI-powered analysis, Pythia, we’ve validated screen hits by measuring cartilage regeneration capacity in patient-derived human cells and discovered three novel therapeutics that exceeded the performance of FGF18, a drug that demonstrated cartilage regenerative properties in phase 2 clinical trials. Notably, both horses and human samples capture some individual differences in response to therapy, which is an important complication in clinical trials.

The predictive power of this experiment was demonstrated to be 80%. While these results are early in the preclinical validation and development process, we couldn’t be more excited about the potential as we move these discoveries forward.

An Asset Discovery Factory

Mosaic Screening has already produced the largest in vivo perturbation dataset on the planet, and we are just getting started.

We are building a vertically integrated pharmaceutical company, with the intention of bringing dozens of blockbuster therapeutics for prevalent age-related disorders into and through clinical trials. We are sitting on more therapeutic data than we can push into development ourselves. We use this data to engage the broader pharma community, leveraging discovery and asset partnerships to cause as many therapies to enter the clinic as possible.

We have developed the world’s first in vivo asset discovery factory, allowing us to generate preclinical revenue while pushing internal development programs into the clinic. In this way, we achieve our mission: curing age-related disease.

Who says money can’t buy time?

We are always looking for excellent team members, mission-aligned investors, and world-class advisors to join us in our journey. Please reach out to us.

Gordian Biotechnology

We’re building a machine that turns therapeutic target hypotheses into more time for people with age-related disease.

By Effie Klimi