Cytotoxic chemotherapies have been a mainstay of cancer treatment for over 40 years. As they are typically administered without the use of precision biomarkers, some patients can suffer severe side effects without any benefit. The development of novel biomarkers to enable precision use of these therapies could reduce toxic side effects, improve overall response rates and reduce unnecessary healthcare expenditures. In this study we use chromosomal instability (CIN) signatures to predict a patient's response to platinum-based and doxorubicin chemotherapy. We retrospectively validated our predictor of platinum sensitivity across 41 high grade serous ovarian cancer patients. Additionally, we discovered a new biomarker for doxorubicin sensitivity based on a CIN signature related to focal DNA amplification, which we retrospectively validated across 26 patients treated with doxorubicin following platinum. We also assessed the performance of these predictors using circulating tumour DNA. As multiple CIN signature biomarkers can be quantified using a single genomic test, this represents a unified approach to guide multiple therapy choices for cytotoxic chemotherapies with the potential to shift current one-size-fits all chemotherapy treatment towards precision medicine.

Using data from electronic medical records we were able to rapidly generate temporal network data. This data can then be loaded into a modern graph database and used to generate a temporal graph of the data. Using a specialist graph language for rapidly querying these graph databases, we are able to rapidly extract temporal path information about patient to patient contact networks based on shared ward encounters. This information can then be used to calculate various network statistics of interest that may be important for clinical use.

Background and Objectives: The burst of high-throughput omics technologies has given rise to a new era in systems biology, offering an unprecedented scenario for deriving meaningful biological knowledge through the integration of different layers of information. Methods: We have developed a new software tool, MOMIC, that guides the user through the application of different analysis on a wide range of omic data, from the independent single-omics analysis to the combination of heterogeneous data at different molecular levels. Results: The proposed pipeline is developed as a collection of Jupyter notebooks, easily editable, reproducible and well documented. It can be modified to accommodate new analysis workflows and data types. It is accessible via momic.us.es, and as a docker project available at github that can be locally installed. Conclusions: MOMIC offers a complete analysis environment for analysing and integrating multi-omics data in a single, easy-to-use platform.