Francesca Vitali <https://deptmedicine.arizona.edu/profile/francesca-vitali-phd>.
Data fusion strategies for precision medicine and drug repurposing.
Data fusion strategies for precision medicine and drug repurposing. Over the last few years, biomedical research and clinical practice have experienced an
incredible growth in terms of both amount and heterogeneity of data being collected and leveraged for different types of analysis. This data explosion represents a great opportunity to increase our knowledge about many biological mechanisms as well as to improve medical processes (i.e., diagnosis, prognosis, therapy). However, not all big data are created equal. The downside of data heterogeneity is it complicates integration analysis. For example, clinical record data is highly heterogeneous, sparsely annotated, and contains several measurement types and unstructured text fields, comprised of ambiguous statements as well as varying levels of certainty, whereas genomic and imaging data are crisp and densely annotated data with a low cardinality of distinct variables. Integrating these data is particularly challenging when the molecular measurements are not conducted on individual subjects. In order to take full advantage of the wide spectrum of biomedical data available, advanced data integration tools need to be applied. In this context, I will present data fusion strategies for precision medicine and drug repositioning from my own research. These methods will include an approach for the prediction of potential multi-target drug repurposing strategies and its performances when applied to triple negative breast cancer. A second method that will be presented computes patient similarities by integrating patient-specific genomic data and public biomedical knowledge through a matrix tri-factorization approach. Finally, I will present a network-based approach integrating genomic and drug data with Gene Ontology-based information theoretic semantic similarities for the suggestion of new drug repurposing candidates. These examples show the potential of developing new research hypotheses and conducting predictive and data interpolation operations.