Thanks to rapid advances in single-cell RNA sequencing (scRNA-seq) technologies, it is now possible to quantify genome-wide gene expression profiles at single-cell resolution at affordable costs. scRNA-seq data hold enormous potential to decipher molecular mechanisms that modulate progression and treatment response in complex diseases such as cancer and to inform personalized treatment regimens. Since scRNA-seq data are extremely sparse and high-dimensional, reducing their dimensionality is a necessary preprocessing step for all analyses. Yet, when applied to scRNA-seq data, standard dimensionality reduction techniques have been shown to yield highly non-canonical and unstable results with unforeseeable effects on downstream analyses.
In NetMap, we will therefore integrate dimensionality reduction for scRNA-seq data with a central task in the detection of molecular disease mechanisms: the identification of key gene regulatory networks and transcription factors that drive cell differentiation. The new computational methods will be applied on scRNA-seq data for exhausted and non-exhausted CD4 T cells, where we will aim at deriving hypotheses on gene regulatory programmes driving T cell exhaustion. The most promising hypotheses will then be tested in pre-clinical in vivo studies.