Current cancer therapies provide targeted treatments attacking specific cells, however, tumor cells are heterogeneous and evolving. To develop personalized treatments, we need to understand the composition of cell types in the tumor and the disrupted regulatory mechanisms that lead to cancer stem cells (CSCs). CSCs are resistant to standard therapies and have the ability to form new tumors leading to relapse and metastasis. Immunotherapies harnessing the immune system can be particularly successful in targeting CSCs, however, their mechanisms of action are not well understood. I hypothesize that an unbiased study of the complex tumor microenvironment containing elusive resistant CSCs and interacting immune populations can be achieved with high-dimensional genome-wide data, such as state-of-the-art single-cell resolution transcriptional integrated with epigenetic measurements, using Bayesian statistical tools that are ideal for distinguishing technical noise from biological heterogeneity and integrating different data types. I capitalize on our previous work in collaboration with the Alexander Rudensky Lab on characterizing immune cell populations in breast cancer tumors, using a computational method we developed in the Dana Pe?er Lab for clustering cells in single-cell transcriptomic data while simultaneously normalizing cells and correcting batch effects. In my PhD work, I showed the power of incorporating epigenetic data in inferring regulatory programs. Hence, in my K99 mentored phase, I aim to develop a computational framework for integrating epigenetic data with single-cell transcriptomic data to infer leukemic stem cells and dysregulated mechanisms in Acute Myeloid Leukemia in collaboration with Ross Levine (Aim 1). I have chosen AML as it involves enrichment of epigenetic mutations and the normal hematopoiesis system is well-characterized and would serve as a reference. As an independent investigator in the R00 phase, I will extend this framework to infer CSCs and dysregulations in the tumor as well as composition of immune cells and their reprogramming in under-characterized solid tumors, in collaboration with Benjamin Neel and others in my future institute (Aim 2). I then aim to use this toolbox to study the impact of immunotherapy treatments on the tumor-immune microenvironment in collaboration with Catherine Wu and my future institute (Aim 3). We expect that our results lead to insights into regulatory mechanisms that are disrupted in cancer and drive heterogeneous populations. We would also infer mechanisms of action of immunotherapies in the tumor-immune microenvironment. This proposal describes a training plan to advance my career to an independent investigator at the interface of machine learning and cancer biology. During the K99 phase, I will be supported by an outstanding and interdisciplinary team of advisors and collaborators with expertise in all aspects of the proposed research. Together with institutional support from Memorial Sloan Kettering Cancer Center and formal coursework and training, I will bridge my knowledge gap in cancer biology and gain the communication and leadership skills vital for my transition.
Cancer treatments lead to favorable outcomes only in a subset of patients due to significant heterogeneity in tumor cells. To develop treatments tailored to each patient and the composition of cell types in each tumor, we need to characterize the complex populations of tumor cells and their dysregulated mechanisms. I propose that an unbiased characterization can be achieved with an interdisciplinary framework of single-cell resolution transcriptional and epigenetic measurements, that are integrated with principled computational tools.
