Shortly after our last release of coronavirus research findings , Westlake University released another breakthrough in COVID-19 research. Tiannan Guo and co-workers identified characteristic molecular changes in the sera from severe #COVID-19 cases, allowing prediction of severe cases using a machine learning model based on serum protein and metabolite biomarkers.

The Guomics Laboratory of Big Proteomic Data, led by Assistant Professor Tiannan Guo, performed the first proteomic and metabolomic characterization of #COVID-19 sera, and managed to identified a series of characteristic biomarkers indicating the severity of COVID-19 patients.

The article is now published online by Cell.

 (Link: https://www.cell.com/cell/fulltext/S0092-8674(20)30627-9）

COVID-19 is an ongoing unprecedented global threat. More than 1,5 million individuals are confirmed cases worldwide and this number is rapidly increasing. Studies investigating the clinical symptoms and epidemiology have been reported, however, little is known about the molecular pathogenesis of SARS-CoV-2, the pathogen of COVID-19. Little clue are available for clinicians to determine why certain patients develop into severe cases, and how to treat them effectively.

In collaboration with clinicians and metabolomics scientists, the team performed a rigorous proteomic and metabolomic analysis of 99 sera samples from four groups including healthy donors, non COVID-19 patients with similar clinical characteristics as COVID-19 patients, non-severe COVID-19 patients, and severe COVID-19 patients. Together, they quantified 894 proteins and 941 metabolites. This study identified characteristic molecules expressed in the blood of severe patients.

(Picture: Experiment design and procedure)

In the blood from severe COVID-19 patients, 93 proteins and 204 metabolites were found to be significantly dysregulated compared to the non-severe COVID-19 cases. Specifically, 50 proteins are involved in pathways including macrophage functions, the complement system and the platelet degranulation. They also found a significant drop of more than 100 amino acids and more than 100 lipids, which may indicate huge consumption of the relevant metabolites during the replication of virus. These findings may provide clues to medical care.

(Picture: The Key proteins and metabolites characterized in severe COVID-19 patients in a working model. Dysregulated molecules involved in macrophage, the complement system and platelet degranulation are shown. These proteins and metabolites are potential biomarkers and therapeutic targets severe cases.)

Furthermore, Guo’s team established a classifier composed of 22 proteins and 7 metabolites using machine learning to identify severe COVID-19 cases. COVID-19 patients who carry such features in their sera exhibited a high risk of getting worse. A model based on the above findings might be used to predict severe cases, facilitating efficiently allocation of medical resources, although further clinical studies may be required to verify it.

Mass spectrometer-based proteomics is a crucial tool for clinical diagnosis and therapeutics.  By integrating data from clinic, proteomics, and metabolomics, the team revealed a holistic landscape of the characteristic molecular changes in the blood of severe COVID-19 cases.

(Picture: Guomics Laboratory of Proteomic Big Data)

Guo’s lab will continue to advance our understanding of COVID-19 pathogenesis process using proteomics technologies, with a goal to improve the diagnosis and treatment of this ongoing disease.

This project is in collaboration with Taizhou Hospital of Zhejiang Province and Calibra Metabolite Lab in Di’an Diagnostics. This research is partly funded by Tencent Foundation.