High-resolution sample size enrichment of single-cell multi-modal low-throughput Patch-seq datasets

Single-cell multimodal technologies are becoming the hot topic of single-cell heterogeneity and function studies, promising to unravel the hidden relationship and functionalities of different aspects of the cells. Among the plethora of single-cell technologies, interesting is the patch-seq technology, which simultaneously performs Patch clamp measures and scRNA-seq on the same cells. However, given the experimental limitations of throughput of Patch clamp, the scRNA-seq analysis is challenging because it requires more samples to investigate cellular heterogeneity. Usually, the solution is associating the cells with the cell types in an existing scRNA-seq dataset. However, doing so loses part of the single cell resolution of the multimodal technique. Therefore, this work proposes a procedure leveraging the Seurat Integration process to find from a reference dataset t he most similar cells to the ones from the patch-seq. The similarity is how much gene expression profiles are identical, and to evaluate that, this work defines various etrics based on R and Index. In this way, one obtains a selection of suitable Reference cells to enrich the number of cells on which to perform multimodal investigation.