Deciphering tumour cell interactions and communications from the gene expression profiles of single cells RNA sequencing data

Author(s): Nicolò Gnoato,Laura Masatti,Stefania Pirrotta,Paolo Martini,Chiara Romualdi,Enrica Calura

Affiliation(s): University of Padova



Cancer is a complex pathological condition that originates from the accumulation of genetic mutations, which can manifest as both point mutations in single nucleotides and structural modifications of the genome, such as copy number variations (CNVs). Lines of evidence have shown that specific genes' copy number variations disrupt their gene expression levels, and the normal cellular physiological mechanisms, triggering uncontrolled growth and division. The formation of tumours results from this abnormal cellular proliferation, which can damage surrounding tissues and impair the body's normal functions. Therefore, characterising CNVs is crucial for investigating the mechanisms underlying tumorigenesis and tumour progression. Today one of the most widely used methods for cell classification relies on marker genes. Of course, these approaches come with a series of limitations due to the selection of the marker genes and the efficiency of the method itself, especially in the context of cancer cells detection. To overcome this problem the primary objective is to develop a new strategy for stratifying normal and tumour cells, and tumoral subclones based on their copy number profiles inferred from single-cell RNA sequencing data (scRNAseq). Since there is evidence that tumour cells can exhibit a significantly more unstable CN profile compared to healthy cells, the idea is to use elements such as 'copy number burden' (CN burden) and 'CNVs signature features” to define a set of scores that effectively distinguish between tumour and non-tumour cells. Moreover, literature indicates that certain genomic regions linked to specific tumour subtypes undergo CN variations. Consequently, our method incorporates this by calculating scores separately for these regions, alongside the entire transcriptome, strengthening in this way, the confidence of our classifications. Given that the tumour microenvironment (TME) plays a key role in tumour progression and recurrence, the second aim of this study is to implement a dedicated automatic pipeline that, starting from the CNV profile specific for each tumoural subcluster, allows to explore how different CNV profiles within tumour cells affect the interaction with the TME. This approach seeks to shed light on potential variations in TME interactions based on CNV profiles, further elucidating the complex dynamics underlying tumorigenesis and tumour progression.