Market analysis of AI-based drug development of biopharmaceuticals : Independent Project work in Molecular Biotechnology

Sammanfattning

Artificial intelligence (AI) is emerging as a transformative tool in biopharmaceutical Research and Development (R&D), including for optimizing complex bioprocess parameters such as pH, temper- ature, dissolved oxygen (DO), and media composition. Traditional process development relies on extensive Design of Experiments (DoE) and iterative trial-and-error experimentation, which is time consuming and resource intensive. This report maps the current landscape of AI solutions available on the market, exploring how AI-driven approaches can accelerate and enhance optimization by leveraging protein sequence and structural data to predict optimal conditions early in development. Our focus areas include sequence-based solubility and expression yield prediction models. It also includes hybrid modeling that combines mechanistic bioprocess knowledge with machine learning, and digital twin simulations of cell culture processes, as well as AI-assisted protein engineering for property improvement. Our findings indicate that AI models can identify critical process parameters and their interactions, enabling more efficient DoE and reducing the experimental workload. For example, virtual bioprocess platforms now allow thousands of condition combinations to be tested in silico within minutes, theoretically pinpointing optimal conditions and cutting required wet-lab experiments dramatically. Similarly, hybrid AI models incorporating fundamental biochemical ki- netics can overcome limited early-stage data, providing interpretable insights into the input-output relationships of the system. The findings demonstrate that integrating AI tools in early bioprocess development can shorten development timelines, increase yields and product quality, and inform better process design decisions. Notably, while these approaches show great promise, the report also discusses current challenges, such as data availability and model interpretability, and outlines how ongoing advances are addressing them, positioning AI as a valuable component in the R&D toolkit for bioprocess optimization.