Our Philosophy of Learning

At Alden Scientific, we believe that progress in science begins with a simple but demanding discipline: learning from failure.

Not failure as an outcome to be avoided, but failure as a source of knowledge and wisdom. Every experiment that does not produce the expected result, every study that cannot be replicated, every sample that behaves unexpectedly, and every hypothesis that proves incomplete contains knowledge. The challenge is not eliminating failure. The challenge is learning from it quickly enough, deeply enough, and systematically enough that it never occurs in the same way twice.

This philosophy emerged from a deeply personal question: how do you know what to do, and how do you know if it worked?

The question arose during founder Jamie Heywood's nine-year journey alongside his brother Stephen, who lived with ALS for more than three times the average survival for the disease. Despite extraordinary efforts, Stephen ultimately died before science could advance quickly enough to deliver a treatment. That experience shaped a conviction that continues to guide our work today: the greatest challenge in medicine is often not intelligence, effort, or even technology. It is time.

How do we shorten the time required to understand disease? How do we accelerate the process of learning what works, what does not, and why?

For more than three decades, our team has pursued those questions through a series of ambitious scientific efforts, each building upon the lessons of the last.

One of the earliest efforts was an unprecedented drug screening program at ALS TDI. Over five years, more than 250 studies evaluated 120 drug candidates representing nearly every actionable target proposed in the ALS literature. The result was both disappointing and illuminating. None produced a meaningful extension of survival in animal models.

Yet the most important lesson was not the absence of a treatment. It was the discovery that many previously reported findings could not be replicated. In seeking a positive control, we uncovered evidence that portions of the scientific literature contained conclusions that could not withstand rigorous testing. This experience reinforced a principle that remains central to our work: literature is an essential starting point, but measurement is the ultimate source of truth.

The lesson was profound. If knowledge cannot be reliably reproduced, it cannot reliably guide action.

The next question became how to generate new knowledge rather than simply test existing ideas. Through some of the first longitudinal studies of state biology in mammals, our team explored how biological systems change over time and which biological signals emerge earliest during disease progression. Rather than treating biology as a static snapshot, these studies approached it as a dynamic process unfolding across time.

This shift—from measuring biology at a moment to measuring biology as a trajectory—became a defining principle of our work.

Yet understanding disease required understanding people, not just animal models.

That realization led to PatientsLikeMe, one of the world's first large-scale platforms for measuring human health experience longitudinally. While often described as a patient social network, PatientsLikeMe was fundamentally a human measurement system. It was designed to understand how symptoms, treatments, outcomes, and lived experiences evolve over time across thousands of individuals.

The platform demonstrated the power of real-world evidence to accelerate learning. In one notable example, PatientsLikeMe showed that lithium, a therapy generating considerable excitement in ALS, provided no measurable benefit to patients. Years later, multiple clinical trials costing hundreds of millions of dollars reached the same conclusion.

The lesson was not simply that lithium failed. The lesson was that better measurement can dramatically shorten the time required to learn.

PatientsLikeMe expanded into multiple sclerosis, Parkinson's disease, mental health, sleep disorders, and many other conditions. It contributed new outcome measures, supported drug development efforts, and improved understanding of disease progression. Yet despite its impact, one challenge remained unresolved. Human experience alone could not fully explain human biology.

To go deeper, we launched DigitalMe, one of the first comprehensive longitudinal human multi-omics studies ever conducted. DigitalMe was designed to measure human biology across every major layer of biological information—genomics, proteomics, metabolomics, immune biology, clinical measurements, behavioral signals, and more—while following individuals over time.

The study represented more than $100 million in investment and generated one of the richest longitudinal biological datasets of its kind. More importantly, it revealed another critical lesson: when measured at sufficient scale, quality, and frequency, proteins encode human physiology in remarkably powerful ways. They provide a dynamic representation of health, disease, aging, and response to intervention that becomes increasingly informative as data quality improves.

At roughly the same time, the scientific community reached an inflection point. Through decades of investment and collaboration, initiatives such as the UK Biobank began producing population-scale proteomic and multi-omic datasets of unprecedented depth and quality. For the first time, biology possessed the scale necessary to begin building predictive models of human physiology.

Alden Scientific emerged from this convergence of ideas, data, and lessons.

The company is not the product of a single discovery. It is the accumulation of decades of learning about what makes biological measurement trustworthy, useful, and ultimately actionable.

Measurement precedes discovery.

Before biology can be modeled, it must be measured consistently. Before datasets can be compared, they must be harmonized. Before predictions can be trusted, sources of variance must be identified and understood.

Much of our work has therefore focused on the practical realities of biological measurement. We learned how samples degrade during transport. We learned how bias enters laboratory systems. We learned how analytical pipelines create unintended distortions. We learned how difficult it is to compare data generated by different technologies, different studies, and different populations.

Each failure became a lesson. Each lesson became a process. Each process became part of a system designed to continuously improve.

As a result, Alden was built around a simple operational principle: never fail the same way twice.

Our laboratory infrastructure, logistics systems, quality controls, harmonization frameworks, and analytical pipelines are all designed to identify sources of variance, understand their causes, and systematically reduce them. This manufacturing-engineering approach to biology enables more reliable measurements and more trustworthy scientific conclusions.

The same philosophy extends into our artificial intelligence systems.

SerenityAI was developed not simply as a tool for analyzing biological data, but as a system for accelerating learning itself. Built upon decades of work in measurement, physiology, and computational biology, SerenityAI helps identify hidden sources of error, reveal previously unseen biological relationships, and generate new hypotheses.

Every day, SerenityAI asks questions that are difficult for humans alone to formulate. Why is this sample behaving differently? What source of variance explains this result? Why does this disease appear to separate into distinct biological subtypes? What can this failure teach us?

In many ways, SerenityAI is a direct extension of the philosophy that created Alden. It exists to help us learn faster.

Combined with advances in modern AI, including large language models and next-generation computational tools, SerenityAI enables a relatively small team to integrate, analyze, and understand biological information at a scale that would have been impossible only a few years ago. It helps transform isolated measurements into connected understanding and accelerates the cycle of observation, learning, and improvement.

Ultimately, our philosophy of learning remains grounded in a deeply human belief.

Every sample belongs to someone.

Every data point represents a person, a family, a hope, a question, or a struggle. The purpose of measurement is not measurement itself. The purpose is to create truthful insights that help people make better decisions about health and disease.

Stephen's disease lasted nine years. It was enough time to teach us many lessons. It was not enough time to help him.

Everything we have built since then—from ALS TDI and PatientsLikeMe to DigitalMe, from large-scale multi-omics to SerenityAI—has been guided by a single objective: reducing the time required to know.

  • To know what is happening in the body.

  • To know whether an intervention is working.

  • To know why one person responds differently than another.

  • To know enough, early enough, to make a difference.

That pursuit continues to define Alden Scientific today.