Development of computational methods to infer cell-cell communication using single-cell RNA sequencing data

In the era of big data, data science and bioinformatics have a pivotal role in analyzing and interpreting complex biological data. This requires the development of advanced mathematical models and efficient algorithms to extract meaningful insights from high-volume data, such as those involved in cell-cell communication. Cell-cell communication is a vital process through which cells constantly send and receive signals with each other to coordinate and regulate numerous biological processes, from maintaining homeostasis and driving cell development to mediating disease progression. Deciphering cellular communication mechanisms is of paramount importance in comprehending the molecular basis of the physiopathology of a living organism. Advancements in next-generation sequencing technologies have revolutionized our understating of molecular biology. In this endeavor, single-cell RNA sequencing technology has emerged as powerful and invaluable tool to reveal cellular heterogeneity in gene expression patterns in a high-throughput manner and at the individual cell level, with a resolution never possible before. This technological breakthrough has led to the development of numerous computational tools designed to systematically infer cellular communication mechanisms from single-cell RNA sequencing data. However, despite the availability of various computational methods, the bioinformatics analysis of cell-cell communication remains a relatively young and rapidly evolving research area, with large room for improvement in the methodological and computational area. The aim of this Ph.D. research project was to design and develop novel computational methods to advance the analysis of cellular communication from single-cell RNA sequencing data. First, a complete characterization of the cellular communication inference landscape was performed, identifying state-of-the-art methods and highlighting the complexity of the biological questions and the diversification of the approaches proposed in literature. In response to key challenges, three computational tools, namely scSeqComm, scSeqCommDiff, and CellGOSSIP, were developed. Each tool addresses specific methodological and computational challenges associated with different aspects of cellular communication, including intercellular and intracellular signaling, as well as differential analysis across conditions. The validation and assessment of these methods demonstrated their robustness and reliability in providing accurate and biologically meaningful results. Overall, this thesis advances the state-of-the-art in the analysis of cell-cell communication, offering novel computational methods that enhance our understanding of the complexity of cellular communication mechanisms in diverse biological contexts.