Proteomics is the large-scale study of the proteome, which consists of all the proteins produced in an organism, cell, or biological system. Using high resolution mass spectrometry, data from the proteome can provide insight into proteins and their signaling pathways as the drivers of human physiology and disease. Yatiri Bio explicitly prioritizes this data as the key to understanding how and why particular drugs successfully eradicate disease. Recent advances in this technology allow for the identification and quantification of thousands (>8,000) of proteins per sample. In addition, the signaling pathways can be accurately monitored via analysis of post translational modifications (PTM) that steer biochemical responses. Consequently, these analyses address the relevant disease on a cellular level including direct observation of the effects of drug perturbation. The advantage of studying the proteome (versus the current standard of identifying only a selective few biomarkers) is that we are able to identify more robust biomarker signatures, improve patient selection, and drive pathway-based therapies for personalized medicine.
Yatiri Bio develops proprietary computational tools (differential expression, pathway enrichment, sample outliers and other analyses) that employ machine learning techniques such as unsupervised clustering and dimensionality-reduction methods to specifically explore proteomics data in the context of human biology. We move beyond the statistical modeling by employing our vast experience and deep understanding of experimental design with the goal of utilizing all elements of the captured data and converting it to actionable knowledge. The result of this work is a fully interactive data portal ProteoBrowserTM and ProteoPathwayTM tailored for each experiment, and optimized with collaborative iterations.
Yatiri Bio has developed a portfolio of cellular models backed by clinical proteome data that are tailored to match molecularly defined cancer subtypes. Our ProteoModelsTM can be used for optimized patient selection, rational exploration of combinational therapies, drug repurposing, identifying resistance mechanisms, and drug discovery with robust and translatable biomarkers from the outset.
While most compounds show robust anti-cancer activity in cellular and animal models, 80% of the compounds that progress to clinical trials fail due to poor correlation between the models and the patients they are meant to represent. The incorporation of accurate efficacy models will streamline the entire drug discovery process, guide early research choices, and make it easier and far less costly to match the right therapeutic with the right patients.