Shaun Jackman

That was a very insightful read! Beyond the Classical Type I Error: Bayesian Metrics for Bayesian Designs Using Informative Priors https://lnkd.in/eFJiwcXB A key reason for the limited use of Bayesian designs in pivotal trials is the challenge of strictly controlling the type I error when borrowing external information, a requirement that regulators often enforce stubbornly. However, dismissing historical information solely to preserve the type I error is a debatable stance. But if some flexibility in Type I error control is allowed, then alternative metrics become necessary to evaluate Bayesian designs. This perspective is supported by the FDA’s guidance on Complex Innovative Trial Desings, which states: “When Type I error probability is not applicable (e.g., some Bayesian designs that borrow external information), appropriate alternative trial characteristics should be considered”. This is precisely the topic addressed in the paper by Best et al., where the authors distinguish between trials that borrow information on the control arm versus those that borrow directly on the treatment effect. That was a very insightful read!

25

1 Comment

🤖 BenchSci & Mila join forces to supercharge AI-driven drug discovery Toronto’s BenchSci and Montréal’s Mila - Quebec Artificial Intelligence Institute are teaming up to push the limits of biological inference with advanced AI models—aiming to cut preclinical timelines and boost research efficiency. 📊 Highlights: – 🔬 AI applied to biological reasoning & hypothesis generation – ⚙️ Integration of deep learning with molecular data – 🤝 Collaboration bridging Canada’s AI & biotech hubs 💡 This partnership exemplifies how Canadian innovation in AI is reshaping the future of drug discovery.

6

We’re excited to see our PROGENITER project, led by Dr. Sriram Subramaniam, featured in Innovation UBC’s latest story that highlights their cross-border partnership with Alloy Therapeutics, Inc. By integrating Alloy’s antibody discovery platforms with PROGENITER’s cryo-EM capabilities, this collaboration will help accelerate the development of antibody therapeutics for infectious disease targets – so that when emerging health threats arise, we can respond faster with life-saving therapies for Canadians. This partnership unites science, technology and open teamwork to position our country as a global leader in immune-based innovation. 🔗 Learn more about PROGENITER: https://lnkd.in/dNmUs5bQ 🔗 Read the news announcement: https://lnkd.in/d5bQyCpb #AntibodyTherapeutics #PandemicPreparedness #Biomanufacturing #LifeSciences https://lnkd.in/g_gX6VyQ

23

Pyxa Has Landed. Today we announce the launch of Pyxa, the first commercial platform for 3D spatial multi-omics in intact, thick tissue. By analyzing tissue up to 100 µm thick, Pyxa moves spatial biology beyond the limitations of 2D thin-tissue sections and allows researchers to see the native organization of cells, molecules, and microenvironments - in three dimensions. This marks a fundamental shift in how researchers understand tissue biology and cellular dynamics. Read more about Pyxa and our limited-time early access promotion: https://hubs.li/Q043QCcd0

233

9 Comments

🧬 Can Canada lead AI-driven drug discovery without scale? In this new Thought Leadership piece for Techcouver, Dylan Roskams-Edris (he/him) and Estrid Jakobsen of Conscience argue that British Columbia’s advantage in life sciences won’t come from outspending or outscaling the U.S. or China — but from collaboration, open science, and shared data infrastructure. BC is already a global leader in lipid nanoparticles, antibodies, and computational drug discovery, with companies like Variational AI, HTuO Biosciences, and Redwood AI pushing the field forward. But fragmented health data, limited capital, and global competition remain structural barriers. The authors make a clear case: ➡️ AI drug discovery is limited less by algorithms than by access to large, high-quality, real-world datasets ➡️ No single Canadian firm can solve this alone — and that’s okay ➡️ Open, pre-competitive collaboration can be a strategic advantage, not a threat to IP They point to real-world examples already underway, including Structural Genomics Consortium, the CACHE Challenges, and BC institutions like Djavad Mowafaghian Centre for Brain Health, Simon Fraser University, and University of British Columbia. The takeaway: Canada doesn’t need to win on scale to win on impact. By leading on standards, data-sharing frameworks, and responsible open science, BC can help shape the future of global drug discovery. #LifeSciences #AI #DrugDiscovery #OpenScience #Biotech #HealthTech https://lnkd.in/geQyTJix

9

I wanted to share some recent work from the Hesperos Inc. team. We had two articles published: a Technology Feature in @NaturePortfolio Lab Animal highlights our Human-on-a-Chip platform alongside other leaders in the field, and a separate publication in @SpringerNature In Vitro Models describes our new sarcopenia-on-a-chip model. The sarcopenia model is a functional system that captures key aspects of age-related muscle loss and fatigue to help support development of better treatments for patients with this disease of high unmet need. Link in the comments to learn more about both. Grateful to be part of this work. #HumanOnAChip #Hesperos #NAMs #Sarcopenia #PreclinicalResearch

23

2 Comments

From student research to global biotech innovation, University of Waterloo alumnus Ryan Denomme (BASc ’10, MASc ’12) is leading the charge.   As CEO and founder of Nicoya Lifesciences, Denomme launched the company through Velocity to solve a major research barrier: traditional tools for analyzing drug interactions were too expensive and slow.   Today, the Kitchener-based company is making a global impact: - Operating in 40+ countries - Creating AI-powered tools for faster drug discovery - Employing co-op students and alumni, including 50% of its executive team Nicoya is helping researchers make faster breakthroughs in health science with its SPR systems, including Alto™, the world’s first digital microfluidics-based model. Behind that innovation is a growing team powered by University of Waterloo talent, like engineering grad Terese Dimeck (BASc ’23), who now works full-time on the company’s flagship product.   More: https://lnkd.in/g2QS2n-c #UWaterloo #VelocityStartups #Engineering #WaterlooAlumni

119

PhyKIT v2.1.8 now supports phylogenetic signal testing - Blomberg's K and Pagel's lambda for continuous traits Method was inspired and validated against phytools across 95 simulated datasets (r = 1.0 for K, r > 0.999 for lambda) Documentation: https://lnkd.in/gdMRNpjQ GitHub: https://lnkd.in/gBV4Mjuh

8