Figure 2.

Overview of the HUSCH workflow and functions. The datasets in the HUSCH database were automatically parsed from multiple different public databases. All datasets were then processed using a uniform workflow based on MAESTRO, including quality control, batch effect removal, cell clustering, differential expression analysis, cell type and tissue level integration. For each dataset, HUSCH displayed relevant study information, including tissue type, assembly, platform, the number of patients and cells, and related study information. In the Dataset Exploration module, HUSCH provides three main functions: single-dataset exploration, single-tissue exploration, and data download. HUSCH also allows online cell type annotation using the curated expression atlas as a reference.