Dermot's Diary #8

The Celtic team has gone big at the 2026 South African Association of Veterinary Technologists Congress. Why the specific focus?
Veterinary molecular diagnostics has become an important part of Celtic's business. 

We've strategically expanded our portfolio and today offer one of the broadest menus of qPCR assays available in Southern Africa, covering both commercial livestock and companion animals. So, it makes sense that we are at SAAVT, and indeed, a sponsor.

What areas of veterinary molecular diagnostics are you involved in?
Our portfolio spans infectious disease diagnostics, genetic testing, and surveillance assays, providing veterinarians, researchers, and agricultural producers with access to world-class molecular diagnostic solutions.

We've also developed an extensive range of veterinary antibodies and research reagents, supporting both diagnostic and scientific applications across the animal health sector.

In a country like ours, your role is vital, right?
Absolutely. Recent outbreaks, such as those of Foot-and-Mouth Disease Virus (FMDV), have highlighted the significant economic impact of animal diseases on agriculture, food security, export markets, and the broader South African economy. Rapid, accurate molecular testing is critical for early detection, disease containment, and effective outbreak management.

Segueing to nowhere, Dermot, talk to us about AI and molecular diagnostics.
Yes, very much on our minds at the moment. The world is trying to grapple with how artificial intelligence (AI) is going to shape things in the very near future, and we're no different.

We previously touched on the US FDA and its approach to regulation, specifically the Predetermined Change Control Plan (PCCP). Can you explain it and whether other global regulators are taking similar approaches?

Sure. The fundamental challenge regulatory bodies face with predictive AI in molecular diagnostics is the shift from static tools to dynamic, continuous-learning systems. Historically, medical device regulations assumed that a product remained unchanged after market approval.

To accommodate machine learning models that update as they process new genomic or clinical data, regulators have shifted towards a Total Product Life Cycle (TPLC) approach. PCCP is a US regulatory mechanism that allows manufacturers to pre-specify planned future modifications to an AI algorithm during its initial approval submission, enabling continuous updates without requiring a new application each time.

And others are following?
Yes. “Scrambling” might be a better descriptor! Most major regulatory bodies are building their frameworks around guidelines from the International Medical Device Regulators Forum (IMDRF). South Africa is actively following the global trend. SAHPRA has formally aligned its regulatory framework with international TPLC principles, borrowing heavily from the FDA and IMDRF guidelines.

More broadly, what major influences will AI have for molecular diagnostics? 

Two main areas, essentially. First, AI accelerates assay design. Machine learning tools are automating and optimising the design of critical components, such as DNA sequences, primers, and crRNAs. The result is high specificity and sensitivity across diverse pathogen variants.

AI also speeds up laboratory workflows. These tools aid data management, automate routine testing, and accelerate the interpretation of complex biological results.

The benefits are incredible, but so are the risks. Hence the enhanced legislation and TPLC principles.