ICTS

Single cell variations within a genetically homogeneous population of cells can lead to significant differences in cell fate in response to external stimuli. This is particularly relevant in cancer cells, where a small population of cells can evade therapies to develop resistance. In my talk, I will present our ongoing work on tracing the origins, nature, and manifestations of single cell variations in response to a variety of cytotoxic chemotherapies and targeted therapies in various cancer models. Our experimental and computational designs promise to provide a foundation for controlling single-cell variabilities in cancer and other biological contexts, such as stem cell reprogramming and transdifferentiation.

Cancer is one of several deadly diseases that affects humankind and its prevalence is increasing alarmingly with time. Understanding cancer biology is getting extremely important not only to discover new and effective therapeutic agents but also to deal with increasing cases of resistance to existing drugs in clinic. There has been many directional efforts to develop and test new-targeted therapies after the availability of cancer genome and frequent and unique targetable mutations found in different types of tumours, however, there are many challenges in this direction to achieve the maximal use of data we have obtained from tumour Omics. Systems biology has a huge potential and it can help in overcoming these challenges and can predict several hidden targets that can be utilised for a better therapeutic outcomes in cancer patients.

Accumulation and selection of somatic mutations in a Darwinian framework result in intra-tumor heterogeneity (ITH) that poses significant challenges to the diagnosis and clinical therapy of cancer. Recently developed single-cell DNA sequencing (SCS) technologies promise to resolve ITH to a single-cell level. However, technical errors in SCS data sets, including false-positives (FP) and false-negatives (FN) due to allelic dropout, and cell doublets, significantly complicate these tasks. While recently developed methods for SNV detection from single-cell DNA sequencing data leverage the evolutionary history of the cells to overcome the technical errors, these methods are not scalable to the extensive genomic breadth of single-cell whole-genome (scWGS) data or the extensive number of cells sequenced using targeted amplicon sequencing approaches.
In this talk, I will describe statistical methods for reconstructing tumor phylogenies from single-cell DNA sequencing datasets. Moreover, I will present scalable mutation calling methods, which extend the phylogeny-guided variant calling approach to sequencing datasets containing millions of loci and thousands of cells respectively. I will present the benchmarking of these methods using simulated datasets with gold-standard ground truth and the application of these methods on human tumor datasets.

Many cells in multicellular organisms change their phenotype. They switch from one cell type to another. Each cell type is a cellular phenotypic state, and switching between cell types is a phenotypic state transition. Epithelial to Mesenchymal Transition (EMT) is an example of such a cellular state transition. Cellular states are usually defined in terms of the expression of certain key molecules, usually known as "molecular markers." Each phenotypic state is a steady state of a dynamical system involving these key molecules. Under this molecular framework, the phenotypic state transition is investigated using the dynamical systems theory. However, molecular markers are proxies to the phenotype of a cell. Certain cellular features, like morphology and motility, are direct phenotypic measures and can be investigated in high-throughput quantitative studies. For example, during EMT, cells undergo distinct morphological changes that could be quantified using microscopy. Can we use the data from quantitative image analysis to understand the dynamical principles of cellular state transition? In this talk, I will address this question. The talk will cover three broad topics - a) Mathematical concepts behind the phenotypic state transition of a cell, b) Experimental and mathematical analysis of morphological state transition in EMT, and c) Characterizing dynamics of the morphological state space during EMT.

Cell fate decisions are made by allowing external signals to govern the steady-state pattern adopted by networks of interacting regulatory factors governing transcription and translation. One of these decisions, of importance for both developmental processes and for cancer metastasis, is the epithelial-mesenchymal transition (EMT). In this talk, we will argue that these biological networks have highly non-generic interaction structures such that they allow for phenotypic states with very low frustration, i.e. where most interactions are satisfied. This property has important consequences for the allowed dynamics of these systems.

Most studies of genetic networks ignore the role played by local reorganization of chromatin structure in determining the dynamics of transcription. However, recent experiments in E coli (related to supercoiling) and cancer cells (related to epigenetic modification of histones) have revealed cases where this is not sufficient. This talk will focus on creating theoretical models which couple small-scale chromatin degrees of freedom to transcriptional dynamics and discuss the consequences for transcriptional noise and for cell fate transitions.

We will discuss (1) the statistical analysis of genome-wide chromosome conformation capture experiments like Hi-C which map the 3D architecture of chromatin and (2) relationships between 3D genomic architecture and gene regulation, including a graph neural network model called GraphReg for predicting gene expression using both 1D genomic/epigenomic data and 3D data. Next we will present a study where dysregulation of 3D architecture in cancer, via a somatic “cohesinopathy”, leads to widespread alternative promoter usage associated with transposable elements (joint work with Ping Chi).

Cancer is characterized by uncontrolled cell growth and proliferation of cells. Epithelial to Mesenchymal Transition (EMT) orchestrated by Transforming Growth Factor-β (TGFβ) signalling plays a crucial role in the invasive and metastatic potential of several carcinomas. EMT is a developmental program which is utilised by cancer metastasis. In this talk, I will discuss the some of the methods that we have employed to develop TGFβ induced epithelial mesenchymal transition map for metastatic breast cancer. The map and the subsequent analyses of the modules identified within the map, offers deeper understanding into the regulatory mechanisms of the complex regulatory networks governing the process of EMT in cancer metastasis.

This is an exciting time for drug development in oncology. I will unroll this talk in 3 parts. First I will talk about how drug development happens today and what are the questions which can be addressed by computational modelers. Next I will talk about the landscape of drug development in oncology today focusing on immuno-oncology, ADCs and bispecifics and the promising results from them that we are seeing.
In the second part of the talk, I will talk about an example of mechanistic thinking informing drug development strategy by highlighting the impact of heterogenous tumor lesions on patient care.

The past twenty-five years have heralded an unparalleled increase in understanding of cancer biology. Over this period, mathematical modelling has emerged as a valuable tool for unravelling the complex processes that contribute to cancer initiation and progression, for testing experimental hypotheses, and assisting with the development of new approaches for improving cancer treatment.

In this lecture, I will focus on modelling approaches that have been used to investigate the growth and response to radiotherapy of solid tumours. These will range from simple, phenomenological models that view the tumour as a homogeneous mass, to more detailed models that resolve the subcellular dynamics of the cell cycle and describe how these dynamics are impacted by fluctuations in oxygen levels, and finally spatially-resolved multiphase models that view the tumour as a heterogeneous mixture of tumour cells, dead cellular material and extracellular fluid and/or resolve different tumour cell phenotypes.

The role of the adaptive immune system in identifying and eliminating cancer populations is now well-appreciated and has ushered in new T cell-based immunotherapies. However, due to the complexity of the adaptive immune system, which involves understanding the behavior of ~109 unique T cell clones in any one individual, nearly all attempts to refine these approaches have been driven by experimental analysis. To predict treatment failure and optimize current therapies, we therefore must understand the dynamic interplay between the adaptive immune system and an evolving cancer population. This talk will focus on our group’s efforts to develop and apply biophysical and stochastic models to gain a quantitative understanding of T cell recognition dynamics and the co-evolution between cancer populations and immune cells.

The therapeutic control of a solid tumour depends critically on the responses of the individual cells that constitute the entire tumour mass. A particular cell's spatial location within the tumour and intracellular interactions, including the evolution of the cell-cycle within each cell, has an impact on their decision to grow and divide. In this session, I will introduce agent based modelling approach (using a cellular automaton) to study the growth and progression of cancer cells. Moreover, in order to address this multiscale nature of solid tumour growth, we will also look at a hybrid, individual-based approach that analyses spatio-temporal dynamics at the level of cells, linking individual cell behaviour with the macroscopic behaviour of cell organisation and the microenvironment.

Despite dramatic progress over the past 50 years, resulting from the synergistic interaction between the Biomedical and Mathematical Sciences, clinical oncology still faces significant hurdles. High Intensity Focused Ultrasound (HIFU) has emerged as a potentially transformative therapy, over the last 20 years. In this talk, I'll touch on current medical obstacles and explore how HIFU offers a new path forward. We'll also delve into the essential role of mathematical and computational modeling in unlocking its potential.

Over the years, mathematical and computational modelling has played a pivotal in understanding the rich dynamics of mammalian cell cycle regulation. In this talk, I would like to highlight some of the methods we have employed in recent times to decode the complex dynamics of mammalian cell cycle regulation under different physiological conditions.

Cell-to-cell variability during TNFα stimulated Tumor Necrosis Factor Receptor 1 (TNFR1) signaling can lead to single-cell level pro- survival and apoptotic responses. This variability stems from the heterogeneity in signal flow through intracellular signaling entities that regulate the balance between these two phenotypes. In this talk, using systematic Boolean dynamic modeling of a TNFR1 signaling network, modulation of the signal flow path variability to enable cells favour apoptosis will be demonstrated. A computationally efficient approach “Boolean Modeling based Prediction of Steady-state probability of Phenotype Reachability (BM- ProSPR)” to accurately predict the network’s ability to settle into different phenotypes will be presented. Model analysis juxtaposed with the experimental observations to unravel the underlying dynamical cross-talk will be discussed.

Next generation DNA sequencing technologies are creating a huge amount of data for organisms across all kingdoms of life. There is need to develop analysis platforms to assist in complex systems analysis. One powerful method combines data on genes, molecules, and proteins with computer modeling. This approach, called Metabolic Genome-Scale Network Reconstruction (GENRE), can help researchers study cells in silico. In this talk, I will go through the principles of how to build a GENRE model and their use to understand fundamental biological processes. This knowledge can lead to new treatments for diseases, ways to create biofuels, and improvements in industrial processes.