ABSTRACT. The immune system is a complex, dynamic and plastic network composed of various interacting cell types that are constantly sensing and responding to environmental cues. From very early on, the immunology field has invested great efforts and ingenuity to characterize the various immune cell types and elucidate their functions. However, accumulating evidence indicates that current technologies and classification schemes are limited in their ability to account for the functional heterogeneity of immune processes. Single cell genomics hold the potential to revolutionize the way we characterize complex immune cell assemblies and study their spatial organization, dynamics, clonal distribution, pathways, and crosstalk. This emerging field can greatly affect basic and translational research of the immune system. I will discuss how recent single cell genomic studies are changing our perspective of various immune related pathologies from cancer to neurodegeneration. Finally, I will consider recent and forthcoming technological and analytical advances in single cell genomics and their potential impact on the future of immunology research and immunotherapy.

ABSTRACT. Recent advances in -omics technologies have provided many valuable insights into complex biological systems. However, the analysis of -omics data is still a challenging analysis frontier, with datasets characterised by high variability, sparsity and technical noise. These complex features make it difficult to discern causual relationships from spurious associations, limiting our ability to obtain novel mechanistic insights, and to optimise the design of resource-intensive downstream experiments. In this study we combine causal network reconstruction, machine learning, and experimental approaches to identify molecular drivers of fate decisions in haematopoietic stem and progenitor cells (HSPCs).

ABSTRACT. Over the last 20 years, network theory applications have shed light onto many important features of a wide range of real-life systems, including complex biological processes. Network modelling allows us to analyse biological systems at different levels, through integrating theoretical and experimental knowledge into a single representation. On the other hand, various mathematical dynamical models have been developed to capture the temporal behaviour for such systems and possibly to highlight new biological insights. While most of the experimental data suffer from sparsity of information, discrete dynamical models have shown to provide a useful qualitative description of the system’s dynamics and behaviour, even in the absence of known kinetic parameters. In this work we focus on network representation and discrete modelling of a gene regulatory network model of monocyte differentiation in normal and tumoural environments. Monocytes are cells in the blood who can produce macrophages, phagocytic cells that can eat pathogens, which are known to display a spectrum of behaviours ranging from an inflammatory phenotype to a fibrotic phenotype, which is used to reconstruct the tissue that was damaged during an inflammatory process. In a tumour setting, some macrophages were found to have an important role protecting cancer cells from killing by lymphocytes (T cells) and their presence confers resistance to chemotherapy (tumour associated macrophages (TAM)s). It is thus important to understand the differentiation process of monocytes into different kinds of macrophages and their impact on tumour development from precancerous lesions to advanced disease. We are working on a model of formation of TAMs in vitro, which reproduces what happens in lymph nodes of Chronic Lymphocytic Leukaemia patients when monocytes enter into contact with large quantities of cancer cells. Here we have focused on two main angles: firstly we use expression data from these in vitro experiments (bulk RNAseq and single cell transcriptomics) and the information from the literature to reconstruct a gene regulatory network of monocyte polarization. The regulatory network shown below will be further elaborated and extended to represent all the macrophage states we see in the experiments. Secondly, we apply logic models to analyse the system’s temporal behaviour under different external conditions, by identifying the attractors in the state space, which will correspond to macrophage phenotypes. The aim of this work is to identify the mechanisms that control the differentiation of monocytes into the various types of macrophages with the prospect to test our results experimentally in controlled in vitro experiments that can be relevant for tumour biology.

ABSTRACT. The application of Systems Biology approaches to Stem Cell research is becoming more and more necessary in order to address a variety of fundamental questions in this field. In particular, with advances in single cell sequencing techniques, it is possible to get a deeper understanding of heterogenous cellular populations. However, the identification of optimal gene sets, whose perturbations can trigger specific cell population shifts, is a challenge in the application of cellular conversion strategies to regenerative medicine. In this talk, I will present computational approaches based on single-cell data, developed in our lab for the identification of optimal cellular reprogramming and differentiation determinants (transcription factors and signaling molecules). In particular, one of the methods, which relies on information theory concepts, aims at identifying synergistic transcriptional identity cores characterizing cell subpopulations. Perturbations of these core transcription factors is shown to induce transitions between cellular subpopulations of heterogeneous stem cell populations. Our computational predictions have been experimentally validated in different cellular systems. Further, I will discuss a recently implemented computational method that integrates cellular signaling and gene regulatory networks to identify key signaling pathways controlled by the niche to maintain specific cellular phenotypes, which could trigger cellular transitions in-vivo. This method can facilitate mimicking the niche effect on stem or progenitor cell states with potential applications in regenerative medicine. In summary, our computational predictions not only have been shown to be useful in guiding experimental research, but are currently being used in designing strategies for cell therapy treatment in patients with Parkinson’s disease, as well as in patients with partial vision loss due to depletion of corneal limbus stem cells.