Section 1.1: The Shift from Medicine to Technology

Q1: As a medical doctor, what prompted your transition from traditional clinical practice to founding Owkin?

A1: My background involved research in lymphoma, where I felt that conventional medicine had reached a plateau. I wanted to introduce a fresh perspective. My exploration into machine learning opened my eyes to its potential to revolutionize our understanding of diseases. This non-linear approach allows for more natural identification of correlations, even if it doesn't fully explain diseases yet. I realized this was my true calling. Additionally, I was eager to embark on an entrepreneurial path.

Section 1.2: Collaborative Foundations

Q2: Was the concept of Owkin solely your brainchild?

A2: Not at all. My goal was to leverage AI to tackle medical challenges and enhance research efforts. I teamed up with a co-founder, an assistant professor in machine learning, and together, we envisioned a platform to create superior drugs, as they are the pinnacle of medical advancement. This led us to develop a federated learning-based data platform.

Subsection 1.2.1: Understanding Federated Learning

Q3: Can you explain federated learning in more detail?

A3: Certainly! We recognized that access to data is crucial for effective machine learning, so we aimed to gather significantly more data than our competitors. Our focus was on obtaining deep, curated data linked to patient samples, with the trust of physicians and patients worldwide. Our unique approach allows us to access high-quality data without owning it, maintaining a collaborative spirit among hospitals and research centers.

Section 2.1: Defining High-Quality Data

Q5: When you mention high-quality data, what exactly do you mean?

A5: Our foundational models are primarily based on histology data, which we complement with genomic data. The synergy between these data types helps us predict specific phenotypes, and we continuously integrate additional modalities to refine our predictions.

Q6: Do you then utilize machine learning on this curated data?

A6: Absolutely. We have a dedicated center to curate and validate data quality, which hospitals can leverage for research. Our machine learning applications aim to uncover new patient subgroups that may respond differently to certain drugs, ultimately enhancing drug discovery.

Section 2.2: Precision Medicine Today and Tomorrow

Q7: How has the precision medicine landscape evolved since Owkin's inception in 2017?

A7: The core concept remains the same: developing targeted treatments for specific patient groups. However, the field is transitioning towards incorporating AI and innovative biomarkers. The future will see a shift from genomic data to include various imaging biomarkers, necessitating the digitization of pathology images.

Q8: So, AI's effectiveness hinges on data digitization?

A8: Exactly. Hospitals need to digitize biopsy data to create a robust pipeline for curated information, paving the way for AI's application in precision medicine.

Q9: Do you think synergy between humans and AI was essential for Owkin's success?

A9: Yes, AI is merely a tool for discovery and communication between data scientists and physicians. Both elements are necessary for success.

Q10: Can you describe the early challenges Owkin faced as a startup?

A10: The journey involved balancing business revenue with partnership development. We also faced the challenge of promoting federated learning, which many were unfamiliar with. Leading an AI biotech is challenging, as the potential of AI in drug development is still being explored.

Section 2.3: Breakthroughs and Future Directions

Q11: When did Owkin experience its breakthrough moment?

A11: Our progress has been gradual, focusing on building a solid foundation without cutting corners. We continually strive for methodical advancement, which has fostered a collaborative environment among our talented team.

Q12: Why did you choose to focus on cancer, and what gaps have you identified?

A12: My background as an oncologist naturally guided our focus, but we also explore cardiovascular diseases now. The wealth of imaging data in oncology presents various avenues for research.

Section 2.4: The Challenge of Neurodegenerative Diseases

Q13: What are your thoughts on precision medicine in neurodegenerative diseases compared to cancer?

A13: Neurodegenerative diseases pose immense challenges due to their complex mechanisms. Unlike cancer, where progress is more evident, advancements in neurology are scarce, making the pursuit even more vital.

Q14: Can you elaborate on your interest in spatial omics?

A14: We are exploring spatial omics to connect genes with tissue at a microenvironmental level, crucial for understanding therapeutic responses. This modality allows us to capture deeper insights that traditional methods may overlook.

Q15: Is it more difficult to study neurodegenerative diseases due to sample accessibility?

A15: Yes, obtaining high-quality samples for neurodegenerative research is challenging, necessitating alternative models and data collection methods.

Q16: What alternatives exist for studying diseases like ALS?

A16: Exploring imaging data and understanding genetic mutations linked to ALS are essential. The field is demanding but incredibly important for addressing this devastating condition.

Section 2.5: Standing Out in a Growing Industry

Q17: As precision medicine evolves, how can companies differentiate themselves?

A17: We view ourselves as an AI biotech rather than a precision medicine company. Our primary goal is to develop innovative drugs that address various patient needs, with precision medicine being a natural offshoot.

Q18: How important is diversity in your team?

A18: Diversity is crucial for fostering urgency and innovative problem-solving in the AI biotech realm.

Q19: If you could redo one decision from the past three years, what would it be?

A19: While there are countless things I might approach differently, I hold no regrets. Each experience has been a learning opportunity that contributed to our growth.

Q20: How do you define success at Owkin?

A20: Success is measured by the number of patients we positively impact. The focus is solely on how our technologies benefit people, not on financial metrics.

Q21: It's clear that your patient-first approach drives Owkin's success.

A21: Absolutely. Prioritizing patient outcomes leads to sound decision-making. Our focus is on making a real difference, regardless of market size.

Q22: Your philosophy of chasing the right goals is refreshing.

A22: Indeed, if you aim to create a meaningful impact, the business will follow.

We extend our gratitude to Thomas Clozel for sharing his insights and to the entire Owkin team for their commitment to transforming cancer treatment.