investigator_user investigator user funding collaborators pending menu bell message arrow_up arrow_down filter layers globe marker add arrow close download edit facebook info linkedin minus plus save share search sort twitter remove user-plus user-minus
  • Project leads
  • Collaborators

Discovery of Synthetic Lethals in Pan-Cancer Genomic Data Using Boolean Methods

Ravindra Majeti

0 Collaborator(s)

Funding source

National Institutes of Health (NIH)
We propose to investigate a computational approach to identify potential therapeutic targets for cancers that harbor specific recurrent mutations usin high-throughput pan-cancer data from primary tumors. Many known mutations that drive cancer are not suitable as direct targets for drug treatment, but the presence of these mutations in some cases creates new dependence on other genes, called synthetic lethals, whose products may consequently be attractive targets for attacking the cancer cells. Existing approaches to find synthetic lethals rely on siRNA or shRNA screens on cell lines, which are laborious, expensive, and often identify synthetic lethals that are not directly clinically applicable becausethey are cell-type or context-dependent. Our approach is to identify a short list of strong candidate synthetic lethal partners for a set of specific recurrent mutations in acute myeloid leukemia (AML) by using pan-cancer Boolean implication mining of primary tumor genome and gene expression data and then functionally validating these candidates. Our overall hypothesis is that, across multiple cancers, synthetic lethal partners of a mutation will be amplified more frequently or deleted less frequently, with concordant changes in expression, in primary tumor samples harboring the mutation of interest. There are two specific aims, one for the computational strategy and one for experimental validation. First, we propose to computationally identify candidate synthetic lethal genes with frequently occurring mutations in AML using Boolean implication mining across multiple human cancers. We will develop a computational method pipeline that can be applied to TCGA mutation, copy number, and expression data to identify candidate synthetic lethal partners of commonly occurring undruggable mutations in AML that also occur in other TCGA cancers: DNMT3A, WT1, and cohesin complex genes. Second, we will employ a two-step validation approach for the top candidate synthetic lethal partners of each mutation in AML. Initially, we will use engineered inducible-mutant cell lines with shRNA lentiviral transduction to assess for synthetic lethality. Those genes validated by thismethod will then be taken forward to final validation in genotyped primary human AML cells using shRNA lentiviral transduction followed by in vitro survival assays and in vivo xenotransplantation assays. We have chosen to focus our synthetic lethality analysis on AML, a bone marrow malignancy with dismal outcome that has not improved for over 3 decades and for which there are no effective targeted therapies. We expect that the proposed study will identify novel druggable targets in AML. Ultimately, the long term objective of this research is to develop a new method for identifying drug targets in many different types of cancer by combining computational and experimental strategies for identifying synthetic lethal partners of recurrent mutations, which we hope will be more efficient and more likely to translate to the clinic than existing approaches.

Related projects