Biotech startup Vivodyne suggests that the AI drug-discovery sector is hindered by a lack of appropriate data. The company has developed modular robotic labs, named HIVE, designed to address this issue.
These HIVE labs are capable of growing 20 types of human tissue, which they then autonomously dose and monitor. This process aims to generate causal biological data, a type of data Vivodyne states is currently missing from AI models and typically sourced from animal testing or studies of single cells or proteins.
Andrei Georgescu, Vivodyne's CEO and co-founder, highlighted that existing AI models lack the data to fully capture the complexity of human biology. He noted that 90% of drugs effective in animal testing fail to gain regulatory approval for human use after entering clinical trials.
Vivodyne, which originated from the University of Pennsylvania in 2021, claims its tissues closely mimic the behaviour of real human organs. For example, its liver cells reportedly have 94% predictive accuracy for toxicity compared to human trials, and its airway tissue matches real human tissue behaviour 96% of the time. Its bone marrow has achieved 100% concordance in tests involving 20 different chemotherapy drugs.
Last week, Vivodyne opened what it describes as the world's largest "human data center" outside San Francisco. The company, which has raised nearly $80 million, states it is collaborating with multiple major pharmaceutical companies.