Fast and Accurate Bacterial Species Identification in Urine Specimens Using LC-MS/MS Mass Spectrometry and Machine Learning

Advances and Progress in Automated Urine Analyzers

The clinical analysis of urine has classically focused on conventional chemical-based urinalysis and urine microscopy. Contemporary advances in both analysis subsets have started to employ new technologies such as automated image analysis, flow cytometry, and mass spectrometry. In addition to new detection technologies, current analyzers have incorporated more advanced imaging, automated sample handing, and machine learning analyses into their workflow. The most advanced semiautomated analyzers can be interfaced with hospital medical record systems, and in the point-of-care setting, smartphones can be used for image analysis. This review will discuss current technological advancements in the field of urinalysis and urine microscopy.

Discrimination of Common E. coli Strains in Urine by Liquid Chromatography-Ion Mobility-Tandem Mass Spectrometry and Machine Learning

Accurate identification of bacterial strains in clinical samples is essential to provide an appropriate antibiotherapy to the patient and reduce the prescription of broad-spectrum antimicrobials, leading to antibiotic resistance. In this study, we utilized the combination of a multidimensional analytical technique, liquid chromatography-ion mobility-tandem mass spectrometry (LC-IM-MS/MS), and machine learning to accurately identify and distinguish 11 Escherichia coli (E. coli) strains in artificially contaminated urine samples. Machine learning was utilized on the LC-IM-MS/MS data of the inoculated urine samples to reveal lipid, metabolite, and peptide isomeric biomarkers for the identification of the bacteria strains. Tandem MS and LC separation proved effective in discriminating diagnostic isomers in the negative ion mode, while IM separation was more effective in resolving conformational biomarkers in the positive ion mode. Using hierarchical clustering, the strains are clustered accurately according to their group highlighting the uniqueness of the discriminating biomarkers to the class of each E. coli strain. These results show the great potential of using LC-IM-MS/MS and machine learning for targeted omics applications to diagnose infectious diseases in various environmental and clinical samples accurately.

Bioinformatics in urology - molecular characterization of pathophysiology and response to treatment

The application of bioinformatics has revolutionized the practice of medicine in the past 20 years. From early studies that uncovered subtypes of cancer to broad efforts spearheaded by the Cancer Genome Atlas initiative, the use of bioinformatics strategies to analyse high-dimensional data has provided unprecedented insights into the molecular basis of disease. In addition to the identification of disease subtypes - which enables risk stratification - informatics analysis has facilitated the identification of novel risk factors and drivers of disease, biomarkers of progression and treatment response, as well as possibilities for drug repurposing or repositioning; moreover, bioinformatics has guided research towards precision and personalized medicine. Implementation of specific computational approaches such as artificial intelligence, machine learning and molecular subtyping has yet to become widespread in urology clinical practice for reasons of cost, disruption of clinical workflow and need for prospective validation of informatics approaches in independent patient cohorts. Solving these challenges might accelerate routine integration of bioinformatics into clinical settings.

Spectral analysis and sorting of microbial organisms using a spectral sorter

This chapter discusses the problems related to the application of conventional flow cytometers to microbiology. To address some of those limitations, the concept of spectral flow cytometry is introduced and the advantages over conventional flow cytometry for bacterial sorting are presented. We demonstrate by using ThermoFisher's Bigfoot spectral sorter where the spectral signatures of different stains for staining bacteria are demonstrated with an example of performing unmixing on spectral datasets. In addition to the Bigfoot's spectral analysis, the special biosafety features of this instrument are discussed. Utilizing these biosafety features, the sorting and patterning at the single cell level is optimized using non-pathogenic bacteria. Finally, the chapter is concluded by presenting a novel, label free, non-destructive, and rapid phenotypic method called Elastic Light Scattering (ELS) technology for identification of the patterned bacterial cells based on their unique colony scatter patterns.

Discovery of Species-Specific Proteotypic Peptides To Establish a Spectral Library Platform for Identification of Nontuberculosis Mycobacteria from Mass Spectrometry-Based Proteomics

Nontuberculous mycobacteria are opportunistic bacteria pulmonary and extra-pulmonary infections in humans that closely resemble Mycobacterium tuberculosis. Although genome sequencing strategies helped determine NTMs, a common assay for the detection of coinfection by multiple NTMs with M. tuberculosis in the primary attempt of diagnosis is still elusive. Such a lack of efficiency leads to delayed therapy, an inappropriate choice of drugs, drug resistance, disease complications, morbidity, and mortality. Although a high-resolution LC-MS/MS-based multiprotein panel assay can be developed due to its specificity and sensitivity, it needs a library of species-specific peptides as a platform. Toward this, we performed an analysis of proteomes of 9 NTM species with more than 20 million peptide spectrum matches gathered from 26 proteome data sets. Our metaproteomic analyses determined 48,172 species-specific proteotypic peptides across 9 NTMs. Notably, M. smegmatis (26,008), M. abscessus (12,442), M. vaccae (6487), M. fortuitum (1623), M. avium subsp. paratuberculosis (844), M. avium subsp. hominissuis (580), and M. marinum (112) displayed >100 species-specific proteotypic peptides. Finally, these peptides and corresponding spectra have been compiled into a spectral library, FASTA, and JSON formats for future reference and validation in clinical cohorts by the biomedical community for further translation.

Recent progress in mass spectrometry-based urinary proteomics

Serum or plasma is frequently utilized in biomedical research; however, its application is impeded by the requirement for invasive sample collection. The non-invasive nature of urine collection makes it an attractive alternative for disease characterization and biomarker discovery. Mass spectrometry-based protein profiling of urine has led to the discovery of several disease-associated biomarkers. Proteomic analysis of urine has not only been applied to disorders of the kidney and urinary bladder but also to conditions affecting distant organs because proteins excreted in the urine originate from multiple organs. This review provides a progress update on urinary proteomics carried out over the past decade. Studies summarized in this review have expanded the catalog of proteins detected in the urine in a variety of clinical conditions. The wide range of applications of urine analysis-from characterizing diseases to discovering predictive, diagnostic and prognostic markers-continues to drive investigations of the urinary proteome.

Deciphering the structure, function, and mechanism of lysine acetyltransferase cGNAT2 in cyanobacteria

Lysine acetylation is a conserved regulatory posttranslational protein modification that is performed by lysine acetyltransferases (KATs). By catalyzing the transfer of acetyl groups to substrate proteins, KATs play critical regulatory roles in all domains of life; however, no KATs have yet been identified in cyanobacteria. Here, we tested all predicted KATs in the cyanobacterium Synechococcus sp. PCC 7002 (Syn7002) and demonstrated that A1596, which we named cyanobacterial Gcn5-related N-acetyltransferase (cGNAT2), can catalyze lysine acetylation in vivo and in vitro. Eight amino acid residues were identified as the key residues in the putative active site of cGNAT2, as indicated by structural simulation and site-directed mutagenesis. The loss of cGNAT2 altered both growth and photosynthetic electron transport in Syn7002. In addition, quantitative analysis of the lysine acetylome identified 548 endogenous substrates of cGNAT2 in Syn7002. We further demonstrated that cGNAT2 can acetylate NAD(P)H dehydrogenase J (NdhJ) in vivo and in vitro, with the inability to acetylate K89 residues, thus decreasing NdhJ activity and affecting both growth and electron transport in Syn7002. In summary, this study identified a KAT in cyanobacteria and revealed that cGNAT2 regulates growth and photosynthesis in Syn7002 through an acetylation-mediated mechanism.

Present and future perspectives on mass spectrometry for clinical microbiology

In the last decade, MALDI-TOF Mass Spectrometry (MALDI-TOF MS) has been introduced and broadly accepted by clinical laboratory laboratories throughout the world as a powerful and efficient tool for rapid microbial identification. During the MALDI-TOF MS process, microbes are identified using either intact cells or cell extracts. The process is rapid, sensitive, and economical in terms of both labor and costs involved. Whilst MALDI-TOF MS is currently the gold-standard, it suffers from several shortcomings such as lack of direct information on antibiotic resistance, poor depth of analysis and insufficient discriminatory power for the distinction of closely related bacterial species or for reliably sub-differentiating isolates to the level of clones or strains. Thus, new approaches targeting proteins and allowing a better characterization of bacterial strains are strongly needed, if possible, on a very short time scale after sample collection in the hospital. Bottom-up proteomics (BUP) is a nice alternative to MALDI-TOF MS, offering the possibility for in-depth proteome analysis. Top-down proteomics (TDP) provides the highest molecular precision in proteomics, allowing the characterization of proteins at the proteoform level. A number of studies have already demonstrated the potential of these techniques in clinical microbiology. In this review, we will discuss the current state-of-the-art of MALDI-TOF MS for the rapid microbial identification and detection of resistance to antibiotics and describe emerging approaches, including bottom-up and top-down proteomics as well as ambient MS technologies.

A fully interpretable machine learning model for increasing the effectiveness of urine screening

OBJECTIVES

This article addresses the need for effective screening methods to identify negative urine samples before urine culture, reducing the workload, cost, and release time of results in the microbiology laboratory. We try to overcome the limitations of current solutions, which are either too simple, limiting effectiveness (1 or 2 parameters), or too complex, limiting interpretation, trust, and real-world implementation ("black box" machine learning models).

METHODS

The study analyzed 15,312 samples from 10,534 patients with clinical features and the Sysmex Uf-1000i automated analyzer data. Decision tree (DT) models with or without lookahead strategy were used, as they offer a transparent set of logical rules that can be easily understood by medical professionals and implemented into automated analyzers.

RESULTS

The best model achieved a sensitivity of 94.5% and classified negative samples based on age, bacteria, mucus, and 2 scattering parameters. The model reduced the workload by an additional 16% compared to the current procedure in the laboratory, with an estimated financial impact of €40,000/y considering 15,000 samples/y. Identified logical rules have a scientific rationale matched to existing knowledge in the literature.

CONCLUSIONS

Overall, this study provides an effective and interpretable screening method for urine culture in microbiology laboratories, using data from the Sysmex UF-1000i automated analyzer. Unlike other machine learning models, our model is interpretable, generating trust and enabling real-world implementation.

BioDiscViz: A visualization support and consensus signature selector for BioDiscML results

Machine learning (ML) algorithms are powerful tools to find complex patterns and biomarker signatures when conventional statistical methods fail to identify them. While the ML field made significant progress, state of the art methodologies to build efficient and non-overfitting models are not always applied in the literature. To this purpose, automatic programs, such as BioDiscML, were designed to identify biomarker signatures and correlated features while escaping overfitting using multiple evaluation strategies, such as cross validation, bootstrapping and repeated holdout. To further improve BioDiscML and reach a broader audience, better visualization support and flexibility in choosing the best models and signatures are needed. Thus, to provide researchers with an easily accessible and usable tool for in depth investigation of the results from BioDiscML outputs, we developed a visual interaction tool called BioDiscViz. This tool provides summaries, tables and graphics, in the form of Principal Component Analysis (PCA) plots, UMAP, t-SNE, heatmaps and boxplots for the best model and the correlated features. Furthermore, this tool also provides visual support to extract a consensus signature from BioDiscML models using a combination of filters. BioDiscViz will be a great visual support for research using ML, hence new opportunities in this field by opening it to a broader community.

Clinical Impact of Rapid Bacterial Microbiological Identification with the MALDI-TOF MS

Rapid microbiological reports to clinicians are related to improved clinical outcomes. We conducted a 3-year quasi-experimental design, specifically a pretest-posttest single group design in a university medical center, to evaluate the clinical impact of rapid microbiological identification information using MALDI-TOF MS on optimizing antibiotic prescription. A total of 363 consecutive hospitalized patients with bacterial infections were evaluated comparing a historical control group (CG) (n = 183), in which the microbiological information (bacterial identification and antibiotic susceptibility) was reported jointly to the clinician between 18:00 h and 22:00 h of the same day and a prospective intervention group (IG) (n = 180); the bacterial identification information was informed to the clinician as soon as it was available between 12:00 h and 14:00 h and the antibiotic susceptibility between 18:00 h and 22:00 h). We observed, in favor of IG, a statistically significant decrease in the information time (11.44 h CG vs. 4.48 h IG (p < 0.01)) from the detection of bacterial growth in the culture medium to the communication of identification. Consequently, the therapeutic optimization was improved by introducing new antibiotics in the 10-24 h time window (p = 0.05) and conversion to oral route (p = 0.01). Additionally, we observed a non-statistically significant decrease in inpatient mortality (global, p = 0.15; infection-related, p = 0.21) without impact on hospital length of stay. In conclusion, the rapid communication of microbiological identification to clinicians reduced reporting time and was associated with early optimization of antibiotic prescribing without worsening clinical outcomes.

Resolving the Temporal Splenic Proteome during Fungal Infection for Discovery of Putative Dual Perspective Biomarker Signatures

Fungal pathogens are emerging threats to global health with the rise of incidence associated with climate change and increased geographical distribution; factors also influencing host susceptibility to infection. Accurate detection and diagnosis of fungal infections is paramount to offer rapid and effective therapeutic options. For improved diagnostics, the discovery and development of protein biomarkers presents a promising avenue; however, this approach requires a priori knowledge of infection hallmarks. To uncover putative novel biomarkers of disease, profiling of the host immune response and pathogen virulence factor production is indispensable. In this study, we use mass-spectrometry-based proteomics to resolve the temporal proteome of Cryptococcus neoformans infection of the spleen following a murine model of infection. Dual perspective proteome profiling defines global remodeling of the host over a time course of infection, confirming activation of immune associated proteins in response to fungal invasion. Conversely, pathogen proteomes detect well-characterized C. neoformans virulence determinants, along with novel mapped patterns of pathogenesis during the progression of disease. Together, our innovative systematic approach confirms immune protection against fungal pathogens and explores the discovery of putative biomarker signatures from complementary biological systems to monitor the presence and progression of cryptococcal disease.

Accurate noise-robust classification of Bacillus species from MALDI-TOF MS spectra using a denoising autoencoder

Bacillus strains are ubiquitous in the environment and are widely used in the microbiological industry as valuable enzyme sources, as well as in agriculture to stimulate plant growth. The Bacillus genus comprises several closely related groups of species. The rapid classification of these remains challenging using existing methods. Techniques based on MALDI-TOF MS data analysis hold significant promise for fast and precise microbial strains classification at both the genus and species levels. In previous work, we proposed a geometric approach to Bacillus strain classification based on mass spectra analysis via the centroid method (CM). One limitation of such methods is the noise in MS spectra. In this study, we used a denoising autoencoder (DAE) to improve bacteria classification accuracy under noisy MS spectra conditions. We employed a denoising autoencoder approach to convert noisy MS spectra into latent variables representing molecular patterns in the original MS data, and the Random Forest method to classify bacterial strains by latent variables. Comparison of the DAE-RF with the CM method using the artificially noisy test samples showed that DAE-RF offers higher noise robustness. Hence, the DAE-RF method could be utilized for noise-robust, fast, and neat classification of Bacillus species according to MALDI-TOF MS data.

Mass spectrometry-based proteomics as an emerging tool in clinical laboratories

Mass spectrometry (MS)-based proteomics have been increasingly implemented in various disciplines of laboratory medicine to identify and quantify biomolecules in a variety of biological specimens. MS-based proteomics is continuously expanding and widely applied in biomarker discovery for early detection, prognosis and markers for treatment response prediction and monitoring. Furthermore, making these advanced tests more accessible and affordable will have the greatest healthcare benefit.This review article highlights the new paradigms MS-based clinical proteomics has created in microbiology laboratories, cancer research and diagnosis of metabolic disorders. The technique is preferred over conventional methods in disease detection and therapy monitoring for its combined advantages in multiplexing capacity, remarkable analytical specificity and sensitivity and low turnaround time.Despite the achievements in the development and adoption of a number of MS-based clinical proteomics practices, more are expected to undergo transition from bench to bedside in the near future. The review provides insights from early trials and recent progresses (mainly covering literature from the NCBI database) in the application of proteomics in clinical laboratories.

Clinical metagenomics-challenges and future prospects

Infections lacking precise diagnosis are often caused by a rare or uncharacterized pathogen, a combination of pathogens, or a known pathogen carrying undocumented or newly acquired genes. Despite medical advances in infectious disease diagnostics, many patients still experience mortality or long-term consequences due to undiagnosed or misdiagnosed infections. Thus, there is a need for an exhaustive and universal diagnostic strategy to reduce the fraction of undocumented infections. Compared to conventional diagnostics, metagenomic next-generation sequencing (mNGS) is a promising, culture-independent sequencing technology that is sensitive to detecting rare, novel, and unexpected pathogens with no preconception. Despite the fact that several studies and case reports have identified the effectiveness of mNGS in improving clinical diagnosis, there are obvious shortcomings in terms of sensitivity, specificity, costs, standardization of bioinformatic pipelines, and interpretation of findings that limit the integration of mNGS into clinical practice. Therefore, physicians must understand the potential benefits and drawbacks of mNGS when applying it to clinical practice. In this review, we will examine the current accomplishments, efficacy, and restrictions of mNGS in relation to conventional diagnostic methods. Furthermore, we will suggest potential approaches to enhance mNGS to its maximum capacity as a clinical diagnostic tool for identifying severe infections.

Proteomics approaches: A review regarding an importance of proteome analyses in understanding the pathogens and diseases

Proteomics is playing an increasingly important role in identifying pathogens, emerging and re-emerging infectious agents, understanding pathogenesis, and diagnosis of diseases. Recently, more advanced and sophisticated proteomics technologies have transformed disease diagnostics and vaccines development. The detection of pathogens is made possible by more accurate and time-constrained technologies, resulting in an early diagnosis. More detailed and comprehensive information regarding the proteome of any noxious agent is made possible by combining mass spectrometry with various gel-based or short-gun proteomics approaches recently. MALDI-ToF has been proved quite useful in identifying and distinguishing bacterial pathogens. Other quantitative approaches are doing their best to investigate bacterial virulent factors, diagnostic markers and vaccine candidates. Proteomics is also helping in the identification of secreted proteins and their virulence-related functions. This review aims to highlight the role of cutting-edge proteomics approaches in better understanding the functional genomics of pathogens. This also underlines the limitations of proteomics in bacterial secretome research.

Robust Capillary- and Micro-Flow Liquid Chromatography-Tandem Mass Spectrometry Methods for High-Throughput Proteome Profiling

Capillary- and micro-flow liquid chromatography-tandem mass spectrometry (capLC-MS/MS and μLC-MS/MS) is becoming a valuable alternative to nano-flow LC-MS/MS due to its high robustness and throughput. The systematic comparison of capLC-MS/MS and μLC-MS/MS systems for global proteome profiling has not been reported yet. Here, the capLC-MS/MS (150 μm i.d. column, 1 μL/min) and μLC-MS/MS (1 mm i.d. column, 50 μL/min) systems were both established based on UltiMate 3000 RSLCnano coupled to an Orbitrap Exploris 240 by integrating with different flowmeters. We evaluated both systems in terms of sensitivity, analysis throughput, separation efficiency, and robustness. capLC-MS/MS was about 10 times more sensitive than μLC-MS/MS at different gradient lengths. Compared with capLC-MS/MS, μLC-MS/MS was able to achieve higher analysis throughput and separation efficiency. During the 7 days' long-term performance test, both systems showed good reproducibility of chromatographic full width (RSD < 3%), retention time (RSD < 0.4%), and protein identification (RSD < 3%). These results demonstrate that capLC-MS/MS is more suitable for high-throughput analysis of clinical samples with a limited starting material. When enough samples are available, μLC-MS/MS is preferred. Together, capLC and μLC coupled to Orbitrap Exploris 240 with moderate sensitivity should well meet the needs of large-cohort clinical proteomic analysis.

Direct Identification of Urinary Tract Pathogens by MALDI-TOF/TOF Analysis and De Novo Peptide Sequencing

For mass spectrometry-based diagnostics of microorganisms, matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) is currently routinely used to identify urinary tract pathogens. However, it requires a lengthy culture step for accurate pathogen identification, and is limited by a relatively small number of available species in peptide spectral libraries (≤3329). Here, we propose a method for pathogen identification that overcomes the above limitations, and utilizes the MALDI-TOF/TOF MS instrument. Tandem mass spectra of the analyzed peptides were obtained by chemically activated fragmentation, which allowed mass spectrometry analysis in negative and positive ion modes. Peptide sequences were elucidated de novo, and aligned with the non-redundant National Center for Biotechnology Information Reference Sequence Database (NCBInr). For data analysis, we developed a custom program package that predicted peptide sequences from the negative and positive MS/MS spectra. The main advantage of this method over a conventional MALDI-TOF MS peptide analysis is identification in less than 24 h without a cultivation step. Compared to the limited identification with peptide spectra libraries, the NCBI database derived from genome sequencing currently contains 20,917 bacterial species, and is constantly expanding. This paper presents an accurate method that is used to identify pathogens grown on agar plates, and those isolated directly from urine samples, with high accuracy.

Rapid Fluorescence Sensor Guided Detection of Urinary Tract Bacterial Infections

Introduction

Urinary tract infections (UTI) are one of the most serious human bacterial infections affecting millions of people every year. Therefore, simple and reliable identification of the urinary tract pathogenic bacteria within a few minutes would be of great significance for diagnosis and treatment of clinical patients with UTIs. In this study, the fluorescence sensor was reported to guide the detection of urinary tract bacterial infections rapidly.

Methods

The Ami-AuNPs-DNAs sensor was fabricated by the amino-modified Au nanoparticles (Ami-AuNPs) and six DNAs signal molecules, which bound to the urinary tract pathogenic bacteria and generated corresponding response signals. Further, based on the collected response signals, identification was performed by principal component analysis (PCA) and linear discriminant analysis (LDA). The Ami-AuNPs and Ami-AuNPs-DNAs were characterized by transmission electron microscopy, UV−vis absorption spectrum, Fourier transform infrared spectrum, dynamic light scattering and zeta potentials. Thereafter, the Ami-AuNPs-DNAs sensor was used to discriminate and identify five kinds of urinary tract pathogenic bacteria. Moreover, the quantitative analysis performance towards individual bacteria at different concentrations were also evaluated.

Results

The Ami-AuNPs-DNAs sensor were synthesized successfully in terms of spherical, well-dispersed and uniform in size, which could well discriminate five main urinary tract pathogenic bacteria with unique fingerprint-like patterns and was sufficiently sensitive to determine individual bacteria with a detection limit to 1×10⁷ cfu/mL. Furthermore, the sensor had also been successfully applied to identify bacteria in urine samples collected from clinical UTIs.

Conclusion

The developed fluorescence sensor could be applied to rapid and accurate discrimination of urinary tract pathogenic bacteria and holds great promise for the diagnosis of the disease caused by bacterial infection.

A Deep-Learning Based System for Rapid Genus Identification of Pathogens under Hyperspectral Microscopic Images

Infectious diseases have always been a major threat to the survival of humanity. Additionally, they bring an enormous economic burden to society. The conventional methods for bacteria identification are expensive, time-consuming and laborious. Therefore, it is of great importance to automatically rapidly identify pathogenic bacteria in a short time. Here, we constructed an AI-assisted system for automating rapid bacteria genus identification, combining the hyperspectral microscopic technology and a deep-learning-based algorithm Buffer Net. After being trained and validated in the self-built dataset, which consists of 11 genera with over 130,000 hyperspectral images, the accuracy of the algorithm could achieve 94.9%, which outperformed 1D-CNN, 2D-CNN and 3D-ResNet. The AI-assisted system we developed has great potential in assisting clinicians in identifying pathogenic bacteria at the single-cell level with high accuracy in a cheap, rapid and automatic way. Since the AI-assisted system can identify the pathogenic genus rapidly (about 30 s per hyperspectral microscopic image) at the single-cell level, it can shorten the time or even eliminate the demand for cultivating. Additionally, the system is user-friendly for novices.

Using Targeted Liquid Chromatography-Tandem Mass Spectrometry to Rapidly Detect β-Lactam, Aminoglycoside, and Fluoroquinolone Resistance Mechanisms in Blood Cultures Growing E. coli or K. pneumoniae

New and rapid antimicrobial susceptibility/resistance testing methods are required for bacteria from positive blood cultures. In this study, a multiplex-targeted liquid chromatography-tandem mass spectrometry (LC-MS/MS) assay was developed and validated for the detection of β-lactam, aminoglycoside, and fluoroquinolone resistance mechanisms in blood cultures growing Escherichia coli or Klebsiella pneumoniae complex. Selected targets were the β-lactamases SHV, TEM, OXA-1-like, CTX-M-1-like, CMY-2-like, chromosomal E. coli AmpC (cAmpC), OXA-48-like, NDM, VIM, and KPC; the aminoglycoside-modifying enzymes AAC(3)-Ia, AAC(3)-II, AAC(3)-IV, AAC(3)-VI, AAC(6′)-Ib, ANT(2′′)-I, and APH(3′)-VI; the 16S-RMTases ArmA, RmtB, RmtC, and RmtF; the quinolone resistance mechanisms QnrA, QnrB, AAC(6′)-Ib-cr; the wildtype quinolone resistance determining region of GyrA; and the E. coli porins OmpC and OmpF. The developed assay was evaluated using 100 prospectively collected positive blood cultures, and 148 negative blood culture samples spiked with isolates previously collected from blood cultures or isolates carrying less prevalent resistance mechanisms. The time to result was approximately 3 h. LC-MS/MS results were compared with whole-genome sequencing and antimicrobial susceptibility testing results. Overall, there was a high agreement between LC-MS/MS results and whole-genome sequencing results. In addition, the majority of susceptible and non-susceptible phenotypes were correctly predicted based on LC-MS/MS results. Exceptions were the predictions for ciprofloxacin and amoxicillin/clavulanic acid that matched with the phenotype in 85.9 and 63.7% of the isolates, respectively. Targeted LC-MS/MS based on parallel reaction monitoring can be applied for the rapid and accurate detection of various resistance mechanisms in blood cultures growing E. coli or K. pneumoniae complex.

Artificial Intelligence in Infection Management in the ICU

This article is one of ten reviews selected from the Annual Update in Intensive Care and Emergency Medicine 2022. Other selected articles can be found online at https://www.biomedcentral.com/collections/annualupdate2022. Further information about the Annual Update in Intensive Care and Emergency Medicine is available from https://link.springer.com/bookseries/8901.

Rapid Bacterial Detection in Urine Using Laser Scattering and Deep Learning Analysis

Images of laser scattering patterns generated by bacteria in urine are promising resources for deep learning. However, floating bacteria in urine produce dynamic scattering patterns and require deep learning of spatial and temporal features. We hypothesized that bacteria with variable bacterial densities and different Gram staining reactions would generate different speckle images. After deep learning of speckle patterns generated by various densities of bacteria in artificial urine, we validated the model in an independent set of clinical urine samples in a tertiary hospital. Even at a low bacterial density cutoff (1,000 CFU/mL), the model achieved a predictive accuracy of 90.9% for positive urine culture. At a cutoff of 50,000 CFU/mL, it showed a better accuracy of 98.5%. The model achieved satisfactory accuracy at both cutoff levels for predicting the Gram staining reaction. Considering only 30 min of analysis, our method appears as a new screening tool for predicting the presence of bacteria before urine culture. IMPORTANCE This study performed deep learning of multiple laser scattering patterns by the bacteria in urine to predict positive urine culture. Conventional urine analyzers have limited performance in identifying bacteria in urine. This novel method showed a satisfactory accuracy taking only 30 min of analysis without conventional urine culture. It was also developed to predict the Gram staining reaction of the bacteria. It can be used as a standalone screening tool for urinary tract infection.

Review of Liquid Chromatography-Mass Spectrometry-Based Proteomic Analyses of Body Fluids to Diagnose Infectious Diseases

Rapid and precise diagnostic methods are required to control emerging infectious diseases effectively. Human body fluids are attractive clinical samples for discovering diagnostic targets because they reflect the clinical statuses of patients and most of them can be obtained with minimally invasive sampling processes. Body fluids are good reservoirs for infectious parasites, bacteria, and viruses. Therefore, recent clinical proteomics methods have focused on body fluids when aiming to discover human- or pathogen-originated diagnostic markers. Cutting-edge liquid chromatography-mass spectrometry (LC-MS)-based proteomics has been applied in this regard; it is considered one of the most sensitive and specific proteomics approaches. Here, the clinical characteristics of each body fluid, recent tandem mass spectroscopy (MS/MS) data-acquisition methods, and applications of body fluids for proteomics regarding infectious diseases (including the coronavirus disease of 2019 [COVID-19]), are summarized and discussed.

DIA Proteomics and Machine Learning for the Fast Identification of Bacterial Species in Biological Samples

Identification of bacterial species in biological samples is essential in many applications. However, the standard methods usually use a time-consuming bacterial culture (24-48 h) and sometimes lack in specificity. To overcome these limitations, we developed a new protocol, combining LC-MS/MS analysis in Data Independent Acquisition mode and machine learning algorithms, enabling the accurate identification of the bacterial species contaminating a sample in a few hours without bacterial culture. In this chapter, we describe the three steps of the protocol (spectral libraries generation, training step, identification step) to generate customized peptide signatures and use them for bacterial identification in biological samples through targeted proteomics analyses and prediction models.

Applications of artificial intelligence in urological setting: a hopeful path to improved care

Artificial intelligence (AI) has been introduced in urology research and practice. Application of AI leads to better accuracy of disease diagnosis and predictive model for monitoring of responses to medical treatments. This mini-review article aims to summarize current applications and development of AI in urology setting, in particular for diagnosis and treatment of urological diseases. This review will introduce that machine learning algorithm-based models will enhance the prediction accuracy for various bladder diseases including interstitial cystitis, bladder cancer, and reproductive urology.

A Novel Classifier Based on Urinary Proteomics for Distinguishing Between Benign and Malignant Ovarian Tumors

Background

Preoperative differentiation of benign and malignant tumor types is critical for providing individualized treatment interventions to improve prognosis of patients with ovarian cancer. High-throughput proteomics analysis of urine samples was performed to identify reliable and non-invasive biomarkers that could effectively discriminate between the two ovarian tumor types.

Methods

In total, 132 urine samples from 73 malignant and 59 benign cases of ovarian carcinoma were divided into C1 (training and test datasets) and C2 (validation dataset) cohorts. Mass spectrometry (MS) data of all samples were acquired in data-independent acquisition (DIA) mode with an Orbitrap mass spectrometer and analyzed using DIA-NN software. The generated classifier was trained with Random Forest algorithm from the training dataset and validated in the test and validation datasets. Serum CA125 and HE4 levels were additionally determined in all patients. Finally, classification accuracy of the classifier, serum CA125 and serum HE4 in all samples were evaluated and plotted via receiver operating characteristic (ROC) analysis.

Results

In total, 2,199 proteins were quantified and 69 identified with differential expression in benign and malignant groups of the C1 cohort. A classifier incorporating five proteins (WFDC2, PTMA, PVRL4, FIBA, and PVRL2) was trained and validated in this study. Evaluation of the performance of the classifier revealed AUC values of 0.970 and 0.952 in the test and validation datasets, respectively. In all 132 patients, AUCs of 0.966, 0.947, and 0.979 were achieved with the classifier, serum CA125, and serum HE4, respectively. Among eight patients with early stage malignancy, 7, 6, and 4 were accurately diagnosed based on classifier, serum CA125, and serum HE4, respectively.

Conclusion

The novel classifier incorporating a urinary protein panel presents a promising non-invasive diagnostic biomarker for classifying benign and malignant ovarian tumors.

Diagnostic Value of the Fimbriae Distribution Pattern in Localization of Urinary Tract Infection

Urinary tract infections (UTIs) are one of the most common infectious diseases. UTIs are mainly caused by uropathogenic Escherichia coli (UPEC), and are either upper or lower according to the infection site. Fimbriae are necessary for UPEC to adhere to the host uroepithelium, and are abundant and diverse in UPEC strains. Although great progress has been made in determining the roles of different types of fimbriae in UPEC colonization, the contributions of multiple fimbriae to site-specific attachment also need to be considered. Therefore, the distribution patterns of 22 fimbrial genes in 90 UPEC strains from patients diagnosed with upper or lower UTIs were analyzed using PCR. The distribution patterns correlated with the infection sites, an XGBoost model with a mean accuracy of 83.33% and a mean area under the curve (AUC) of the receiver operating characteristic (ROC) of 0.92 demonstrated that fimbrial gene distribution patterns could predict the localization of upper and lower UTIs.

Fast detection of E. coli with a novel fluorescent biosensor based on a FRET system between UCNPs and GO@Fe3O4 in urine specimens

Biosensors based on nanomaterials are becoming a research hotspot for the rapid detection of pathogenic bacteria. Herein, a "turn-on" fluorescent biosensor based on a FRET system was constructed for the fast detection of a representative pathogenic microorganism, namely, E. coli, which causes most urinary tract infections. This biosensor was constructed by utilizing synthesized UCNPs as fluorescent donors with stable luminescence performance in complex biological samples and GO@Fe3O4 as a receptor with both excellent adsorption ability and fluorescence quenching ability. A specific ssDNA selected as an aptamer which could recognize E. coli was immobilized on the UCNPs to form UCNP-Apt nanoprobes. The nanoprobes were adsorbed on the surface of GO@Fe3O4 through the π-stacking interactions between aptamers and GO. In the presence of E. coli, UCNP-Apt nanoprobes detached from GO@Fe3O4 due to the specific recognition of aptamers and bacteria, resulting in obvious fluorescence recovery, and the concentration of bacteria was positively correlated with the intensity of the fluorescence signal; such a "turn-on" signal output mode ensures excellent precision. In addition, the easy magnetic separation of GO@Fe3O4 simplifies the operation process, helping the sensor detect bacteria in 30 minutes with a linear range from 103 to 107 CFU mL-1 and a limit of detection of 467 CFU mL-1. Moreover, recovery test results also showed that the sensor has clinical application potential for the rapid detection of pathogenic microorganisms in complex biological samples.

Current applications of artificial intelligence combined with urine detection in disease diagnosis and treatment

In recent years, the advantages of artificial intelligence (AI) in data processing and model analysis have emerged in the medical field, enabled by computer technology developments and the integration of multiple disciplines. The application of AI in the medical field has gradually deepened and broadened. Among them, the development of clinical medicine intelligent decision-making is the fastest. The advantage of clinical medicine intelligent decision-making is to make the diagnosis faster and more accurate on the basis of certain information. Urine detection technologies, such as urine proteomics, urine metabolomics, and urine RNomics, have developed rapidly with the advancements in omics and medical tests. Advances in urine testing have made it possible to obtain a wealth of information from easily accessible urine. However, it has always been a problem to extract effective information from this information and use it. AI technology provides the possibility to process and use the information in urine. AI, combined with urine detection, not only provides new possibilities for precise and individual diagnosis and disease treatment, but also helps promote non-invasive diagnosis and treatment. This article reviews the research and applications of AI combined with urine detection for disease diagnosis and treatment and discusses its existing problems and future development.

Emerging Options for the Diagnosis of Bacterial Infections and the Characterization of Antimicrobial Resistance

Precise and rapid identification and characterization of pathogens and antimicrobial resistance patterns are critical for the adequate treatment of infections, which represent an increasing problem in intensive care medicine. The current situation remains far from satisfactory in terms of turnaround times and overall efficacy. Application of an ineffective antimicrobial agent or the unnecessary use of broad-spectrum antibiotics worsens the patient prognosis and further accelerates the generation of resistant mutants. Here, we provide an overview that includes an evaluation and comparison of existing tools used to diagnose bacterial infections, together with a consideration of the underlying molecular principles and technologies. Special emphasis is placed on emerging developments that may lead to significant improvements in point of care detection and diagnosis of multi-resistant pathogens, and new directions that may be used to guide antibiotic therapy.

The new microbiology: cultivating the future of microbiome-directed medicine

The discovery of human-associated microscopic life forms has captivated the scientific community since their first documentation in the 17th century. Subsequent isolation and cultivation of microorganisms have spurred great leaps in medicine, including the discovery of antibiotics, identifying pathogens that cause infectious diseases, and vaccine development. The realization that there is a vast discrepancy between the number of microscopic cell counts and how many could thrive in the laboratory motivated the advent of sequencing-based approaches to characterize the uncultured fraction of the microbiota, leading to an unprecedented view into their composition and putative function on all bodily surfaces. It soon became apparent that specific members of the microbiota can be our commensal partners with new implications on various aspects of health, as well as a rich source of therapeutic compounds and tools for biotechnology. Harnessing the immense repertoire of microbial properties, however, inadvertently requires pure cultures for validation and manipulation of candidate genes, proteins, or metabolic pathways, just as mammalian cell culture has become an indispensable tool for mechanistic understanding of host biology. Yet, this renewed interest in growing microorganisms, individually or as a consortium, is stalled by the laborious nature of conventional cultivation methods. Addressing this unmet need through implementation of improved media design and new cultivation techniques is arguably instrumental to future milestones in translational microbiome research.

Recent evolutions of machine learning applications in clinical laboratory medicine

Machine learning (ML) is gaining increased interest in clinical laboratory medicine, mainly triggered by the decreased cost of generating and storing data using laboratory automation and computational power, and the widespread accessibility of open source tools. Nevertheless, only a handful of ML-based products are currently commercially available for routine clinical laboratory practice. In this review, we start with an introduction to ML by providing an overview of the ML landscape, its general workflow, and the most commonly used algorithms for clinical laboratory applications. Furthermore, we aim to illustrate recent evolutions (2018 to mid-2020) of the techniques used in the clinical laboratory setting and discuss the associated challenges and opportunities. In the field of clinical chemistry, the reviewed applications of ML algorithms include quality review of lab results, automated urine sediment analysis, disease or outcome prediction from routine laboratory parameters, and interpretation of complex biochemical data. In the hematology subdiscipline, we discuss the concepts of automated blood film reporting and malaria diagnosis. At last, we handle a broad range of clinical microbiology applications, such as the reduction of diagnostic workload by laboratory automation, the detection and identification of clinically relevant microorganisms, and the detection of antimicrobial resistance.