Abstract
RNA viruses have high mutation frequency, quick generation periods and vast population numbers, which promote fast evolution and host environment adaptation. We integrated scRNA-seq and spatial transcriptomics to profile immune cells and viral gene expression in COVID-19. Cell types and interactions were identified using Seurat-based tools. Spatial transcriptomics analysis revealed viral hotspots, and GISAID data were used to track SARS-CoV-2 mutations. Single-cell and spatial transcriptomics analyses revealed that immune cells such as Neutrophils, Monocyte:CD14 + , and T cell:CD4+ central memory are highly abundant in COVID-19 patients, particularly in mild and severe cases, and are concentrated in the central and upper regions of lung tissue. Pseudotime and CellChat analyses indicated that cell differentiation trajectories and communication networks shift toward heightened inflammatory responses in severe conditions. Spatial analysis of viral gene expression showed that SARS-CoV-2 genes, especially Nucleoprotein, Spike and Envelope were highly expressed in central and upper-right tissue regions, suggesting active viral replication. This localized viral activity was strongly associated with areas of immune cell infiltration and inflammation. The top 10 sustainable mutants in SARS-CoV-2 genome with high frequency were observed in NSP12 (P323L, 99%, Switzerland), Spike (D614G, 97%, Switzerland), NSP4 (T492I, 79%, Switzerland), NSP6 (T77A, 70%, Guangdong), Orf9c (G50N, 64%, England), Nucleoprotein (D377Y, 62%, United States), Orf9b (T60A, 61%, France), NSP14 (I42V, 55%, United States), Envelope (T9I, 51.3%, Trinidad and Tobago), and NSP5 (P132H, 51.2%, United States). Following to this approach is crucial for a strong epidemiological reaction against the changing SARS-CoV-2 outbreak.