The incorporation of machine learning methods into proteomics workflows improves the identification of disease-relevant biomarkers and biological pathways. However, machine learning models, such as deep neural networks, typically suffer from lack of interpretability. Here, we present a deep learning approach to combine biological pathway analysis and biomarker identification to increase the interpretability of proteomics experiments. Our approach integrates a priori knowledge of the relationships between proteins and biological pathways and biological processes into sparse neural networks to create biologically informed neural networks. We employ these networks to differentiate between clinical subphenotypes of septic acute kidney injury and COVID-19, as well as acute respiratory distress syndrome of different aetiologies. To gain biological insight into the complex syndromes, we utilize feature attribution-methods to introspect the networks for the identification of proteins and pathways important for distinguishing between subtypes. The algorithms are implemented in a freely available open source Python-package (https://github.com/InfectionMedicineProteomics/BINN).

ApoE is a well-known lipid-binding protein that plays a main role in the metabolism and transport of lipids. More recently, apoE-derived peptides have been shown to exert antimicrobial effects. Here, we investigated the antibacterial activity of apoE using in vitro assays, advanced imaging techniques, and in vivo mouse models. The formation of macromolecular complexes of apoE and endotoxins from Gram-negative bacteria was explored using gel shift assays, transmission electron microscopy, and CD spectroscopy followed by calculation of the α-helical content. The binding affinity of apoE to endotoxins was also confirmed by fluorescent spectroscopy detecting the quenching and shifting of tryptophan intrinsic fluorescence. We showed that apoE exhibits antibacterial activity particularly against Gram-negative bacteria such as Pseudomonas aeruginosa and Escherichia coli. ApoE protein folding was affected by binding of bacterial endotoxin components such as lipopolysaccharide (LPS) and lipid A, yielding similar increases in the apoE α-helical content. Moreover, high-molecular-weight complexes of apoE were formed in the presence of LPS, but not to the same extent as with lipid A. Together, our results demonstrate the ability of apoE to kill Gram-negative bacteria, interact with their endotoxins, which leads to the structural changes in apoE and the formation of aggregate-like complexes.

Recent advances in mass spectrometry-based peptidomics have catalyzed the identification and quantification of thousands of endogenous peptides across diverse biological systems. However, the vast peptidomic landscape generated by proteolytic processing poses several challenges for downstream analyses and limits the comparability of clinical samples. Here, we present an algorithm that aggregates peptides into peptide clusters, reducing the dimensionality of peptidomics data, improving the definition of protease cut sites, enhancing inter-sample comparability, and enabling the implementation of large-scale data analysis methods akin to those employed in other omics fields. We showcase the algorithm by performing large-scale quantitative analysis of wound fluid peptidomes of highly defined porcine wound infections and human clinical non-healing wounds. This revealed signature phenotype-specific peptide regions and proteolytic activity at the earliest stages of bacterial colonization. We validated the method on the urinary peptidome of type 1 diabetics which revealed potential subgroups and improved classification accuracy.

Data independent acquisition mass spectrometry (DIA-MS) has recently emerged as an important method for the identification of blood-based biomarkers. However, the large search space required to identify novel biomarkers from the plasma proteome can introduce a high rate of false positives that compromise the accuracy of false discovery rates (FDR) using existing validation methods. We developed a generalized precursor scoring (GPS) method trained on 2.75 million precursors that can confidently control FDR while increasing the number of identified proteins in DIA-MS independent of the search space. We demonstrate how GPS can generalize to new data, increase protein identification rates, and increase the overall quantitative accuracy. Finally, we apply GPS to the identification of blood-based biomarkers and identify a panel of proteins that are highly accurate in discriminating between subphenotypes of septic acute kidney injury from undepleted plasma to showcase the utility of GPS in discovery DIA-MS proteomics.

The application of machine learning algorithms to facilitate the understanding of changes in proteome states has emerged as a promising methodology in proteomics research. Unfortunately, these methods can prove difficult to interpret, as it may not be immediately obvious how models reach their predictions. We present the data processing kitchen sink (DPKS) which provides reproducible access to classic statistical methods and advanced explainable machine learning algorithms to build highly accurate and fully interpretable predictive models. In DPKS, explainable machine learning methods are used to calculate the importance of each protein towards the prediction of a model for a particular proteome state. The calculated importance of each protein can enable the identification of proteins that drive phenotypic change in a data-driven manner while classic techniques rely on arbitrary cutoffs that may exclude important features from consideration. DPKS is a free and open source Python package available at https://github.com/InfectionMedicineProteomics/DPKS .

Antibodies are critical to the host’s immune defense against bacterial pathogens. Understanding the mechanisms of antibody–antigen interactions is essential for developing new targeted immunotherapies. Building computational workflows that can identify where an antibody binds its cognate antigen and deconvoluting the interaction interface in a high-throughput manner are critical for advancing this field. Cross-linking mass spectrometry (XL-MS) integrated with structural modeling offers a flexible and high-resolution strategy to map protein–protein interactions from low sample amounts. However, cross-linking and in silico modeling have limitations that require robust analytical workflows to make accurate inferences. In this study, we introduce Sidewinder, a modular high-throughput pipeline combining state-of-the-art computational structural prediction and molecular docking with rapid XL-MS analysis, enabling comprehensive interrogation of antibody–antigen systems. We validated this pipeline on antibodies targeting two Streptococcus pyogenes virulence factors. Using recently published data, we identified a well-defined monoclonal antibody epitope on Streptolysin O by generating and querying a large ensemble of interaction models probabilistically. We also showcased the utility of the Sidewinder pipeline by analyzing a more complex system, involving monoclonal antibodies that target the cell wall-anchored M1 protein. The flexibility and robustness of the Sidewinder pipeline provide a powerful framework for future studies of complex antibody–antigen systems, potentially leading to new therapeutic strategies.