ABSTRACT | Next-generation sequencing and other modern biotechnological methods generate vast amounts of data every day. However, extracting relevant information from these sources can be challenging, both with AI-driven and classical statistical methods, due to the black-box phenomena or the difficulty in reaching statistical significance in large datasets.
At Turbine, we simulate cellular decision-making using a manually curated protein-protein interaction-based signaling network, customized with high-throughput in vitro and in vivo biological datasets such as whole-genome sequencing and RNAseq. We then train the network's behavior to recapitulate experimentally observed genetic and pharmacological perturbation response data.
Our Simulated Cell platform is currently being used to identify novel drug targets, biomarkers, and synergistic combinations in hemato-oncology, feeding our internal drug discovery pipeline or external partnerships with pharma companies. In this lecture, we will discuss how we make different data layers work together properly, fine-tune the model to match available cellular responses, and run and translate our own in-silico screens to generate results that meet specific project goals.
Turbine AI is a startup in Budapest that simulates the biology of cells to identify potential drug targets speeding up drug development.
BIO | Ivan graduated from Semmelweis University as a medical doctor. During his time at university, he developed an interest in modeling complex diseases, driven by a desire to tailor therapies to the needs of individual patients rather than the current population-level focus. His early work as an undergraduate focused on network-based modeling of cellular signaling, which eventually led to him co-founding Turbine. Currently, he is working on identifying and validating novel drug targets and biomarkers for internal programs and partnerships, as well as continuously improving the Simulated Cell technology at Turbine.