CellTools algorithm for mapping scRNA-seq query cells to the reference dataset improves the classification of resilient and susceptible retinal ganglion cell types

The clustering of single cell RNA-sequencing (scRNA-seq) data enables the classification of cell types, and the development of integration mapping algorithms has enabled the tracing of altered-from-baseline transcriptomes to their respective cell type origins. Here, we developed an algorithm that removes sources of noise from scRNA-seq reference dataset, and in the next step optimizes weight-assignment to anchors, cumulatively improving the accuracy of query cells mapping to reference dataset. The denoising step of our algorithm also improved the performance of other mapping algorithms. To further demonstrate biological relevance, using our algorithm we determined the type-origin of the 17% of injured retinal ganglion cells (RGCs) that a prior algorithm did not identify. As we found that most of the originally unassigned cells belonged to only some RGC types, a consequent change in the proportions of the surviving types resulted in an amended ranking of resiliency to injury. We also identified new cluster-markers for RGC types, validated two novel markers by immunostaining in retinas, and developed a website for cluster-by-cluster comparison of gene expression between uninjured and injured RGC types. Additional bioinformatic analyses contributed new insights into the global characteristics of RGC types and how axonal injury affects them, showing how dissimilarity between transcriptomes of RGC types increases during maturation and after injury. We further characterized the correspondence between the neonatal and adult RGC types, and showed which cluster markers change expression developmentally or after injury. We also show, for the first time, that global properties of the transcriptome can predict the resilience to injury of at least some cell types. The R-package, CellTools, for the algorithms we developed, will assist scRNA-seq studies across biological fields.