Efficient Integration of Human Skin Single-Cell RNA Sequencing Data

This document presents a comprehensive effort to create an integrated human skin cell atlas, aiming to establish a consensus nomenclature for normal human skin. The research involved reanalyzing 24 scRNA-seq datasets, developing a molecular and spatial atlas of skin fibroblasts, and investigating the role of miR-149 in skin inflammation. A significant achievement was the integration of 463,503 cells from 145 donors across 24 studies, forming the basis for the human cell atlas version 1. Additionally, a novel deep learning-based method, scSEA, was developed to enable unbiased reference mapping and the discovery of new cell types. This work provides a valuable resource for understanding skin biology and disease.