RESUMEN
Background: Classification of disease severity is crucial for the management of COVID-19. Several studies have shown that individual proteins can be used to classify the severity of COVID-19. Here, we aimed to investigate whether integrating the four types of protein context data, namely, protein complexes, stoichiometric ratios, pathways and network degrees will improve the severity classification of COVID-19. Methods: A SWATH-based proteomic data set of 54 sera samples from 40 COVID-19 patients was employed as the training cohort. Results: Machine learning prioritized two complexes, one stoichiometric ratio, five pathways, twelve proteins and five network degrees. A model based on these 25 features led to effective classification of severe cases with an AUC of 0.965, outperforming the models with proteins only. Complement component C9, transthyretin (TTR) and TTR-RBP complex, the stoichiometric ratio of SAA2/ YLPM1, and the network extent of SIRT7 and A2M were highlighted in this classifier. This classifier was further validated with a TMT-based proteomic data set from the same cohort and an independent SWATH-based proteomic data set from Germany, reaching an AUC of 0.900 and 0.908, respectively. Machine learning models integrating protein context information achieved higher AUCs than models with only one feature type. Conclusion: Our results show that the integration of protein context including protein complexes, stoichiometric ratios, pathways, network degrees, and proteins improves phenotype prediction.