The next breakthrough in RNA therapeutics won't come from a single innovation. It requires combining two key elements: a 'programmable' mRNA payload designed to be active only in specific cells, and a targeted delivery system to get it there. This two-part solution represents the next generation of in-vivo therapies.
While AI enables rapid drug creation for single individuals (n-of-1), the economic model is broken. It is not a commercial opportunity, creating an urgent societal challenge to develop new funding mechanisms like public-private partnerships to support these life-saving, non-scalable treatments.
Early AI drug discovery platforms built robust models but often failed to generate relevant outputs. Their lack of deep biological understanding led to flawed data collection and training sets, creating a "garbage in, garbage out" problem where models were disconnected from real-world biology.
For patients with ultra-rare diseases, traditional drug development is too slow. AI platforms like Therna's can design a custom RNA molecule in days and complete the lab-testing cycle in under three months, compressing a multi-year process and making previously impossible treatments viable.
Many scientists are driven by pure curiosity. However, the mindset that pushes an academic toward entrepreneurship is a relentless focus on reaching a definitive conclusion—a 'yes or no' answer. This goal-oriented drive to translate a concept into a real-world application is a key founder trait in biotech.
Building biologically relevant AI is not a one-off process. It demands a continuous "lab in the loop" system where wet lab experiments generate proprietary data to train models, whose outputs are then physically tested in the lab. This iterative feedback cycle constantly refines the model's predictive accuracy.