Rise of the Machine

July 17, 2024 by Kyle Studey

Artificial intelligence and therapeutics

In 2022, the last time Boston hosted the TIDES USA event, the idea of equating Moore’s Law to the manufacture of oligonucleotides and mRNA was but a small ripple amongst the crashing mRNA waves. After all, what did the number of transistors on integrated circuits have to do with therapeutics?

How times change. In 2024, a tenet of Moore’s Law, that gains are made due to experience in production, moved into the spotlight at TIDES. The second keynote session from esteemed Harvard geneticist Dr. George Church set a tone that carried throughout the meeting: machine learning and artificial intelligence (AI) as a key cornerstone of innovation across all areas of RNA therapeutics.

Church’s talk was a broad scope look at the work of several different companies to leverage machine learning, multiplexing, or both, to advance their fields; Dyno Therapeutics is changing the way we design gene vectors and capsid proteins using machine learning, Shape Therapeutics looks to utilize its extensive artificial intelligence-guided libraries to develop next generation AAVs and RNA editing therapeutics, and Nabla Bio combines its AI-guided design process with practical wet lab approaches to design drug candidates.

Closer to the mRNA therapeutic world, innovation is in full swing as well, also guided by AI and machine learning techniques. We are beginning to see the validation of machine learning algorithms used to design optimal mRNA sequences for stability, expression, and localization to desired tissues.

One particularly fascinating session was hosted not by a biologist, but by leading computer science professor Dr. Liang Huang, who boldly proclaimed his platform developed a COVID-19 vaccine transcript sequence that significantly outperformed Pfizer/BioNTech’s sequence in stability, expression, and in vivo antibody response.

It is not just mRNA drug developers evaluating AI to innovate, but mRNA raw material suppliers as well, as the field begins to understand how we can continue to utilize novel modified nucleotides, unnatural cap analogs, and other mRNA modifications for meaningful potency benefits.

As we move beyond the accepted co-transcriptional capping reagents and N1-Methylpseudo-UTP nucleotides, it will be interesting to see the approach drug developers take to selecting and utilizing these innovations; will just a few of the more than 150 developed modifications position themselves at the top across all indications and modalities, or will we see a “toolbox” approach adapted, as specific modifications are characterized and selected according to their ability to modulate transcript expression, durability, and delivery?

It would be wise for mRNA manufacturing partners to evaluate their own toolboxes and consider any nuances that these modifications may introduce to manufacturing approaches, so they are ready to catch the latest technology thrown to them.

TIDES 2024 showed us that we have reached a stride within the mRNA therapeutic industry, getting more comfortable with evaluating novel approaches and raw materials to design, develop, and manufacture more potent mRNA transcripts. And at each of these innovation junctures, AI and machine learning are taking center stage. Wherever the future of this industry takes us, these technologies will be a critical part of taking us there.

• Learn more about Aldevron’s mRNA platform and services
• Have questions on this topic? Contact Kyle Studey
• Have an idea for a topic? Let us know!

Subscribe to our blog