IDENTIFYING AND EXPLOITING SYNTHETIC LETHALITY FOR CANCER THERAPEUTICS

Synthetic Lethality (SL) is considered a promising approach for the development of personalized treatments. However, SL interactions in humans remain largely unknown and there is a need to discover more such pairs, especially those that are clinically actionable, and to develop tools that can enable us to gain deeper insights into their role in cancer and therapy development. In this thesis, new methods have been developed to discover SL pairs and to utilize SL pairs in biological network mining and clinical decision support. Two new methods are designed to identify potential SL pairs using multiple genomic data sources, based on statistical hypothesis testing and collective matrix factorization. A new algorithm to mine and analyse a network of pairwise interactions is presented, finally, an application in personalized treatment recommendations is explored that uses information on SL pairs and other public genomic databases by integrating them within an integer linear programming framework.

Author dc:creator

• HERTY LIANY

• synthetic lethality, cancer therapeutics, maximum weight bipartite subgraphs, collective matrix factorization, personalized drug recommendations, BPM