Many of the challenges in drug discovery and development could be ameliorated if only we could predict earlier in the process which drugs are likely to work and for which patients. At insitro, we are enabling better predictions throughout the pharmaceutical value chain.
Population Scale Data
Our predictive models are grounded in human data. We start by deriving insights from genetic, phenotypic, and clinical data, using modern machine learning to increase the power of traditional genetic analyses by moving closer to the underlying architecture and biology of disease. These genetic insights are a critical first step towards building truly predictive models of disease that capture the causal biology.
Cell-Based Disease Models
Armed with an understanding of disease architecture, we combine patient-derived, induced pluripotent stem cells (iPSCs), genome editing, high content cellular phenotyping, and machine learning to build in vitro models of disease. We design these disease models to be maximally predictive of human clinical outcomes by optimizing for genetics, cell-type, environment, and multidimensional data collection.
Biology at Scale
Powerful machine learning requires powerful data. Our data pipelines and automation infrastructure allow us to go beyond artisanal chemistry and biology, and rapidly generate massive amounts of high-quality data. This scale allows us to span much more of the diversity of human disease and potential therapies.
Cutting-Edge Machine Learning
With massive amounts of high quality data, we then develop and deploy a variety of leading-edge machine learning methods. As witnessed in other industries, machine learning can make sense of vast amounts of high-dimensional data that are beyond human ability to interpret. Our machine learning models allow us to differentiate between cell states at much finer granularity and predict disease-relevant clinical traits.
Predictive Insights and New Medicines
The outcome is an integrated model of disease spanning in vitro cellular systems and in silico machine learning models — an insitro model. Our models can discover previously unseen disease subtypes and search for interventions that move them from an “unhealthy” to a “healthy” state. We couple these models with our team’s extensive experience and expertise in advancing new medicines to identify disease-modifying targets, to enable the drug design process, and to drive new insights on biomarkers and clinical development strategies.