Algorithmic approaches to systems biology, data integration, and evolution

National Library Of Medicine

Investigators

Linked publications, trials & patents

Abstract

Przytycka’s group (AlgoCSB – Algorithms for Computational and Systems Biology research group) develops computational algorithms advancing systems and population-level understanding of cellular functions in health and diseases. Our objective is to provide innovative algorithmic solutions to emerging problems in computational biology and to use these approaches to further the understanding of biomolecular systems with the long-term goal of informing treatment decisions. Currently much of our studies are focused on the research related to tumor evolution with a particular interest in the role of the tumor micro and macro environment and clonal competition. We are the first to utilize stochastic Ornstein-Uhlenbeck (OU) process to model clonal gene expression evolution in tumor. Applying our model to sublines derived from single cells of a mouse melanoma revealed that sublines with distinct phenotypes are underlined by different patterns of gene expression adaptation, indicating non-genetic mechanisms of cancer evolution. The results if these studies are reported in Cell Systems [1]. Following the identification of phenotypically different clones in melanoma tumors, we focused on understanding interactions between them and the tumor microenvironment. To address this challenge, together with our experimental collaborators, we developed dedicated in vivo mouse models and use experimental measurements afforded by these models to develop mathematical models of clonal interactions. Based on the results of the experimental study, we built a Lotka–Volterra-based model interaction between clones and extend it with a piecewise equation for T cell dynamics to capture immune–tumor interactions. By identifying key subclone–immune interactions, this study is expected to provide insights into how intratumor heterogeneity and immune pressure influence melanoma progression and suggests strategies for optimizing therapeutic timing and adaptive treatment. The results of these studies are prepared for publication. Tumor clonal heterogeneity and clonal evolution is also a subject of collaborative paper with Dr. S Cenk Sahinalp: “A Partition Function Algorithm to Evaluate Inferred Subclonal Structures in Single-Cell Sequencing Data” selected for presentation at RECOMB 2025, and currently prepared for publication. In addition a follow-up paper “Improved Algorithms for Bi-partition Function Computation” selected for presentation at WABI 2025 and is also currently prepared for journal publication. We also continue to work on mutational signatures in cancer. Cancer genomes accumulate a large number of somatic mutations resulting from various endogenous and exogenous causes, including normal DNA damage and repair, cancer-related aberrations of the DNA maintenance machinery, and mutations triggered by carcinogenic exposures. Different mutagenic processes lead to different patterns of somatic mutations called mutational signatures. We can think of mutational signatures as fingerprints of the corresponding mutagenic processes acting on a given genome. One of important computational challenges is the identification of signatures acting in a given genome when the data is sparse. In our collaborative studies with Dr. Roded Sharan “Mutational Signature Refitting on Sparse Pan-Cancer Data” accepted for presentation at WABI 2025 and currently prepared for journal publication we developed a machine supervised learning predictor for this task. We also contributed to the paper “APOBEC affects tumor evolution and age at onset of lung cancer in smokers” from Tere Landi group (NCI) published recently in Nature Communications [2]. We are currently working on developing a new signature identification method. Our collaboration with Brian Oliver at Indiana University continues to focus on studies on single cell analysis of sexual dimorphism in fly. We have recently published the paper "Cell-Type Specific Variation in X-Chromosome Dosage Compensation in Drosophila" in microPublication Biology [3]. In this work, leveraging the single-nucleus Fly Cell Atlas (FCA) dataset, which includes 388,918 nuclei across diverse tissues, we investigated cell-type-specific patterns of X-chromosome dosage compensation. We identified a continuum of compensation groups based on their X-to-autosome (X/A) expression ratios ranging from anti-compensated to effectively compensated and overcompensated. Anti-compensation was predominantly observed in male reproductive tissues, while overcompensation was prevalent in neural cells. The expression levels of the dosage compensation machinery's non-coding RNAs, RoX1 and RoX2 , correlated with compensation levels, but were insufficient to fully explain the observed patterns of compensation. These findings reveal the complexity of dosage compensation and suggest that its regulation by the RoX RNAs is nonlinear, implicating potential alternative mechanisms in certain cell types. We also made significant progress on our study of fly allometry. The results of these studies are now summarized in a bioRxiv paper “Cell type specific allometry controls sex-differences in Drosophila body size”. Following well recognized fact that Drosophila females are larger than males, we used Fly Cell Atlas single cell data and dedicated experiments we show that allometry in files involves both: differences in cell sizes and cell numbers. As a new direction, we are currently working on developing and validating a new mathematical method that links spatial transcriptomics features to patient phenotypes