Mass spectrometry combined with machine learning identifies novel protein signatures as demonstrated with multisystem inflammatory syndrome in children

Abstract

Rapid and accurate diagnosis of emerging inflammatory illnesses is challenging due to overlapping clinical features with existing conditions. We demonstrate an approach that integrates proteomic analysis with machine learning to identify diagnostic protein signatures, using the example of SARS-CoV-2-induced multisystem inflammatory syndrome in children (MIS-C). We used plasma samples collected from subjects diagnosed with MIS-C and compared them first to controls with asymptomatic/mild SARS-CoV-2 infection and then to controls with pneumonia or Kawasaki disease. We used mass spectrometry to identify proteins and support vector machine (SVM) algorithm-based classification schemes to identify protein signatures. Diagnostic accuracy was assessed by calculating sensitivity, specificity, and area under the ROC curve (AUC), and corrected for overfitting by cross-validation. Proteomic analysis of a training dataset containing MIS-C (N = 17), and asymptomatic/mild SARS-CoV-2 infected control samples (N = 20) identified 643 proteins, of which 101 were differentially expressed. Plasma proteins associated with inflammation increased, and those associated with metabolism and coagulation decreased in MIS-C relative to controls. The SVM machine learning algorithm identified a three-protein model (ORM1, AZGP1, SERPINA3) that achieved 90.0% specificity, 88.2% sensitivity, and 93.5% AUC, distinguishing MIS-C from controls in the training set. Performance was retained in the validation dataset utilizing MIS-C (N = 19) and asymptomatic/mild SARS-CoV-2 infected control samples (N = 10) (90.0% specificity, 84.2% sensitivity, 87.4% AUC). We next replicated our approach to compare MIS-C with similarly presenting syndromes, such as pneumonia (N = 17) and Kawasaki disease (N = 13), and found a distinct three-protein signature (VWF, FCGBP, and SERPINA3) that accurately distinguished MIS-C from the other conditions (97.5% specificity, 89.5% sensitivity, 95.6% AUC). A software tool was also developed that may be used to evaluate other protein signatures using our data. These results demonstrate that the use of mass spectrometry to identify candidate plasma proteins followed by machine learning, specifically SVM, is an efficient strategy for identifying and evaluating biomarker signatures for disease classification.

Keywords: Biomarkers; Hyperinflammatory illnesses; Long COVID; Support vector machine.

Fig. 1

Support vector machine (SVM) model building and cross-validation. The figure shows the steps of SVM model development used in this work and the sample sets utilized at each step.

Fig. 2

Volcano plot of differentially abundant proteins between MIS-C and mild/asymptomatic SARS-CoV-2. Proteins in the volcano plot are shown as red circles (Holm corrected p value ≤ 0.05), green circles (Holm corrected p value ≤ 0.05 and ≥ twofold change) and orange circles (Holm corrected p value > 0.05 and ≥ twofold change).

Fig. 3

Pathway enrichment analysis of differentially abundant proteins. Pathway analysis was conducted on lists of proteins differentially expressed in MIS-C vs. mild/asymptomatic SARS-CoV-2 infection (FDR < 0.05 and p value < 0.05) and the reactome and gene ontology (GO) biological processes databases. The top 20 pathways ranked by p value are shown for differentially increased proteins in MIS-C (panel A, Reactome; panel B, GO) and differentially decreased proteins in MIS-C (panel C, Reactome; panel D, GO).

Fig. 4

External validation of an SVM model. The figure shows a receiver operating characteristic (ROC) curve visualizing the performance of three proteins (ORM1, SERPINA3, and AZGP1) applied to the validation dataset (MIS-C vs. mild/asymptomatic SARS-CoV-2).

Fig. 5

Differentially abundant proteins between MIS-C, pneumonia, Kawasaki Disease, and mild/asymptomatic SARS-CoV-2. Volcano plots of differentially abundant proteins between (A) MIS-C and pneumonia, (B) MIS-C and Kawasaki disease, and (C) MIS-C and mild/asymptomatic SARS-CoV-2 infection. Green and red circles are proteins having Holm corrected p value ≤  0.05. Green and orange circles are proteins with ≥ twofold change. (D) UpSet plot showing the shared proteins between all pairwise comparisons. MK MIS-C vs. Kawasaki disease, MP MIS-C vs. pneumonia, MA MIS-C vs. mild/asymptomatic SARS-CoV-2 infection.

Fig. 6

Multi-disease comparison and SVM model. (A) Volcano plot of differentially abundant proteins between MIS-C and Kawasaki disease (K), pneumonia (P), and mild/asymptomatic SARS-CoV-2 (M/A). Proteins are depicted as green, red, or orange circles, based on Holm adjusted p value and fold-change, as in Fig. 2. (B) ROC curve visualizing the performance of a 3-protein signature (VWF, FCGBP, and SERPINA3). K Kawasaki disease, P pneumonia, M/A mild/asymptomatic SARS-CoV-2 infection.