Determination of the neighbouring cell types that regulate the gene expression of a single cell has now been made possible, thanks to a statistical method developed by researchers from the University of Tsukuba.
In an important development for the field of bioinformatics, researchers from the University of Tsukuba have developed a statistical method that helps to identify and filter out the biologically relevant cell-cell communications from a vast pool of data regarding spatial gene expression. Their findings were published in Bioinformatics.
Cell-cell communication regulates gene expression essential for the normal function of the cell, as well as disease progression. Presently, although single-cell RNA sequencing and spatially resolved transcriptomics allow one to glean insights into cell-cell communications, some major limitations exist during data analysis.
“Most existing statistical analysis methods do not account for the spatial organisation of cells within an organ composed of various cell types,” states Associate Professor Haruka Ozaki, senior author of the study. He highlights that the location of the cells, the number of the cells, as well as the cell types in the vicinity do affect gene expression in neighbouring cells in reality, thus their interactions have to be accounted for.
To properly account for the above complications in data analysis, the team, created a statistical method, named Cell-Cell communications analysis by Partial Least Square regression modelling (CCPLS), which was able to analyse spatial gene expression data at a single-cell level. The main goal of the method was to recognise and quantitatively measure the influence of neighbouring cell types on the individual cells’ variability in gene expression.
The researchers first used CCPLS on computer-generated data set to verify its ability to accurately recognise and quantify the influence of neighbouring cell types on a cell’s variability in gene expression. They then used the method on a real-life data set and found that their method correctly identified that astrocytes promote the differentiation of oligodendrocyte precursor cells into oligodendrocytes, which was in line with previous research done in mouse models.
In further research, CCPLS was applied to another dataset regarding nine different cell types originating from the colon. Analysis using this statistical method resulted in the unprecedented discovery that the development of immature B cells is influenced by communication with IgA B cells.
Since CCPLS’s ability to identify gene expression variability involving cell-cell communication far exceeded that of current statistical methods, there is a high possibility that it would be of great use to data analytics in the field of bioinformatics, especially in the search for novel therapeutic targets and discovering new effects of cell arrangements on gene expression of various types of cells.
Source: Tsuchiya et al. (2022). CCPLS reveals cell-type-specific spatial dependence of transcriptomes in single cells. bioRxiv.