Purpose: The molecular pathology of lung injury in patients with COVID-19 is still unclear. In this study, we performed a proteomic study of lung tissues from seven patients with COVID-19, and eight controls. Experimental design: Lung parenchymal tissue in COVID-19 was obtained from autopsy samples. Control tissue was obtained from cancer resection samples. Protein was extracted from tissue by phenol extraction. A tandem mass tag-based quantitative proteomic approach combined with bioinformatics analysis was used to detect proteomic changes in the SARS-CoV-2-infected lung tissues. Results: A total of 6602, and 6549 proteins were identified in replicate 1 and 2, respectively. Of which 6332 and 6248 were creditable proteins with Score Sequest HT > 0, unique peptide ≥ 1, and TMT quantifying information. Of them, the differentially expressed proteins were 297, and XXX, respectively. These proteins are enriched in 241 pathways. The upregulated proteins are mainly included neutrophil extracellular trap formation (n = 17), lysosome (n = 12), and phagosome (n = 12). The down-regulated proteins are involved in Tight junction (n = 6), Focal adhesion (n = 6), and Lipid and atherosclerosis (n = 5). Our results systematically outlined the molecular pathological features in terms of the lung response to SARS-CoV-2 infection, and provided the scientific basis for the therapeutic target that is urgently needed to control the COVID-19 pandemic. Our foundational dataset elucidates the biological impact of severe SARS-CoV-2 infection across the body, a key step towards new treatments. Overall, we emphasize the importance of histopathological patient stratifcation when interpreting molecular COVID-19 data.