Teresa Przytycka, PhD

About

Research Interests

The research in my group focuses on developing computational methods for systems biology including application to cancer research and gene regulation and methods for analysis of new types of experimental data.

In the context of cancer, we explore the interplay between co-occurrence and mutual exclusivity of genetic perturbations in the context of functional interactions between genes for predicting cancer related pathways, drug response and study cancer heterogeneity. We also study mutational signatures in cancer. Specifically, mutations in cancer are thought to arise from a combination of errors in DNA replication and from DNA damage caused by external mutagenic factor such as UV light. Various mutagenic processes often produce characteristic mutational patterns called mutational signatures. We are interested in uncovering the interactions between mutational signatures, molecular properties of cancer cells, and exogenous mutagenic factors.

In the context gene regulation our group is particularly interested in predicting context / tissue specific gene regulatory networks. In addition, recognizing that gene regulation is more than a binary relation between TF and target genes we study the role of DNA shape for TF binding and the role of non-canonical DNA structures such as quadruplex, hairpin, Z-DNA and R-loops in gene regulation and mutagenesis. Finally, to advance methods to identify DNA features relevant for gene regulation, we are working on developing biologically interpretable deep learning methods to study translation and other processes involved in gene expression.

Evolutionary analysis of gene expression can shed additional light on context specific gene regulation. Currently, studies of species evolution typically focus on the evolution of species' genetic code, the DNA molecule. However DNA genetic code is the same for all cells in the organism and provides limited information about tissue specific evolutionary processes. To fill this gap, we are developing methods to model gene expression evolution based on a stochastic process and use it to study tissue specific gene expression evolution.

Gene expression across tissues and condition is increasingly measured using single cell resolution technologies. This provides unprecedented opportunity to uncover and to study the variety of cell types present in individual tissues and organs. However this new technology also calls for developing computational approaches allowing for a comparative analysis of such data. To address this need, we recently introduced a computational method, scPopCorn (single-cell subpopulations comparison). To achieve superior performance, we utilized several modern algorithmic techniques such as Google's personalized PageRank approach and graph k-partition (Wang et al., Cell Systems 2019).

Over the last few years we also worked on developing methods for analysis of high throughput HT-SELEX data and for utilizing RNA aptamers for drug design. In particular we developed AptaSUITE: A Full-Featured Bioinformatics Framework for the Comprehensive Analysis of Aptamers from HT-SELEX Experiments (Hoinka et al., Mol Therapy Nucleic Acids. 2018 ) and AptaBlocks: Designing RNA complexes and accelerating RNA-based drug delivery systems (Wang et al., NAR 2018). The software developed by our group is freely accessible to the scientific community as well as to the industry. As of October 2019, AptaSUITE has been uploaded more than 1000 times.

Publications

More