Rapid diagnostics that enable identification of infectious agents improve patient outcomes, antimicrobial stewardship, and length of hospital stay. Current methods for pathogen detection in the clinical laboratory include biological culture, nucleic acid amplification, ribosomal protein characterization, and genome sequencing. Pathogen identification from single colonies by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) analysis of high abundance proteins is gaining popularity in clinical laboratories. Here, we present a novel and complementary approach that utilizes essential microbial glycolipids as chemical fingerprints for identification of individual bacterial species. Gram-positive and negative bacterial glycolipids were extracted using a single optimized protocol. Extracts of the clinically significant ESKAPE pathogens: E nterococcus faecium, S taphylococcus aureus, K lebsiella pneumoniae, A cinetobacter baumannii, P seudomonas aeruginosa, and E nterobacter spp. were analyzed by MALDI-TOF-MS in negative ion mode to obtain glycolipid mass spectra. A library of glycolipid mass spectra from 50 microbial entries was developed that allowed bacterial speciation of the ESKAPE pathogens, as well as identification of pathogens directly from blood bottles without culture on solid medium and determination of antimicrobial peptide resistance. These results demonstrate that bacterial glycolipid mass spectra represent chemical barcodes that identify pathogens, potentially providing a useful alternative to existing diagnostics.

No abstract

Proteomics studies rely on the accurate assignment of peptides to the acquired tandem mass spectra—a task where machine learning algorithms have proven invaluable. We describe mokapot, which provides a flexible semisupervised learning algorithm that allows for highly customized analyses. We demonstrate some of the unique features of mokapot by improving the detection of RNA-cross-linked peptides from an analysis of RNA-binding proteins and increasing the consistency of peptide detection in a single-cell proteomics study.

Exosomes are microvesicles of endocytic origin constitutively released by multiple cell types into the extracellular environment. With evidence that exosomes can be detected in the blood of patients with various malignancies, the development of a platform that uses exosomes as a diagnostic tool has been proposed. However, it has been difficult to truly define the exosome proteome due to the challenge of discerning contaminant proteins that may be identified via mass spectrometry using various exosome enrichment strategies. To better define the exosome proteome in breast cancer, we incorporated a combination of Tandem-Mass-Tag (TMT) quantitative proteomics approach and Support Vector Machine (SVM) cluster analysis of three conditioned media derived fractions corresponding to a 10 000g cellular debris pellet, a 100 000g crude exosome pellet, and an Optiprep enriched exosome pellet. The quantitative analysis identified 2 179 proteins in all three fractions, with known exosomal cargo proteins displaying at least a 2-fold enrichment in the exosome fraction based on the TMT protein ratios. Employing SVM cluster analysis allowed for the classification 251 proteins as "true" exosomal cargo proteins. This study provides a robust and vigorous framework for the future development of using exosomes as a potential multiprotein marker phenotyping tool that could be useful in breast cancer diagnosis and monitoring disease progression.

Infectious diseases have a substantial global health impact. Clinicians need rapid and accurate diagnoses of infections to direct patient treatment and improve antibiotic stewardship. Current technologies employed in routine diagnostics are based on bacterial culture followed by morphological trait differentiation and biochemical testing, which can be time-consuming and labor-intensive. With advances in mass spectrometry (MS) for clinical diagnostics, the U.S. Food and Drug Administration has approved two microbial identification platforms based on matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) MS analysis of microbial proteins. We recently reported a novel and complementary approach by comparing MALDI-TOF mass spectra of microbial membrane lipid fingerprints to identify ESKAPE pathogens. However, this lipid-based approach used a sample preparation method that required more than a working day from sample collection to identification. Here, we report a new method that extracts lipids efficiently and rapidly from microbial membranes using an aqueous sodium acetate (SA) buffer that can be used to identify clinically relevant Gram-positive and -negative pathogens and fungal species in less than an hour. The SA method also has the ability to differentiate antibiotic-susceptible and antibiotic-resistant strains, directly identify microbes from biological specimens, and detect multiple pathogens in a mixed sample. These results should have positive implications for the manner in which bacteria and fungi are identified in general hospital settings and intensive care units.

Tandem mass spectrometry is the only high-throughput method for analyzing the protein content of complex biological samples and is thus the primary technology driving the growth of the field of proteomics. A key outstanding challenge in this field involves identifying the sequence of amino acids -- the peptide -- responsible for generating each observed spectrum, without making use of prior knowledge in the form of a peptide sequence database. Although various machine learning methods have been developed to address this de novo sequencing problem, challenges that arise when modeling tandem mass spectra have led to complex models that combine multiple neural networks and post-processing steps. We propose a simple yet powerful method for de novo peptide sequencing, Casanovo, that uses a transformer framework to map directly from a sequence of observed peaks (a mass spectrum) to a sequence of amino acids (a peptide). Our experiments show that Casanovo achieves state-of-the-art performance on a benchmark dataset using a standard cross-species evaluation framework which involves testing with out-of-distribution samples, i.e., spectra with never-before-seen peptide labels. Casanovo not only achieves superior performance but does so at a fraction of the model complexity and inference time required by other methods.

A fundamental challenge for any mass spectrometry-based proteomics experiment is the identification of the peptide that generated each acquired tandem mass spectrum. Although approaches that leverage known peptide sequence databases are widely used and effective for well-characterized model organisms, such methods cannot detect unexpected peptides and can be impractical or impossible to apply in some settings. Thus, the ability to assign peptide sequences to the acquired tandem mass spectra without prior information -- de novo peptide sequencing -- is valuable for gaining biological insights for tasks including antibody sequencing, immunopeptidomics, and metaproteomics. Although many methods have been developed to address this de novo sequencing problem, it remains an outstanding challenge, in part due to the difficulty of modeling the irregular data structure of tandem mass spectra. Here, we describe Casanovo, a machine learning model that uses a transformer neural network architecture to translate the sequence of peaks in a tandem mass spectrum into the sequence of amino acids that comprise the generating peptide. We train a Casanovo model from 30 million labeled spectra and demonstrate that the model outperforms several state-of-the-art methods on a cross-species benchmark dataset. We also develop a version of Casanovo that is fine-tuned for non-enzymatic peptides. Finally, we demonstrate that Casanovo's superior performance improves the analysis of immunopeptideomics and metaproteomics experiments and allows us to delve deeper into the dark proteome.

Solid-phase peptide synthesis has been applied to the preparation of phosphonate-derivatized oligoproline assemblies containing two different Ru(II) polypyridyl chromophores coupled via "click" chemistry. In water or methanol the assembly adopts the polyproline II (PPII) helical structure, which brings the chromophores into close contact. Excitation of the assembly on ZrO2 at the outer Ru(II) in 0.1 M HClO4 at 25 °C is followed by rapid, efficient intra-assembly energy transfer to the inner Ru(II) (k(EnT) = 3.0 × 10(7) s(-1), implying 96% relative efficiency). The comparable energy transfer rate constants in solution and on nanocrystalline ZrO2 suggest that the PPII structure is retained when bound to ZrO2. On nanocrystalline films of TiO2, excitation at the inner Ru(II) is followed by rapid, efficient injection into TiO2. Excitation of the outer Ru(II) is followed by rapid intra-assembly energy transfer and then by electron injection. The oligoproline/click chemistry approach holds great promise for the preparation of interfacial assemblies for energy conversion based on a family of assemblies having controlled compositions and distances between key functional groups.