Helix Matrix Transformation Combined With Convolutional Neural Network Algorithm for Matrix-Assisted Laser Desorption Ionization-Time of Flight Mass Spectrometry-Based Bacterial Identification.
ABSTRACT: Matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) analysis is a rapid and reliable method for bacterial identification. Classification algorithms, as a critical part of the MALDI-TOF MS analysis approach, have been developed using both traditional algorithms and machine learning algorithms. In this study, a method that combined helix matrix transformation with a convolutional neural network (CNN) algorithm was presented for bacterial identification. A total of 14 bacterial species including 58 strains were selected to create an in-house MALDI-TOF MS spectrum dataset. The 1D array-type MALDI-TOF MS spectrum data were transformed through a helix matrix transformation into matrix-type data, which was fitted during the CNN training. Through the parameter optimization, the threshold for binarization was set as 16 and the final size of a matrix-type data was set as 25 × 25 to obtain a clean dataset with a small size. A CNN model with three convolutional layers was well trained using the dataset to predict bacterial species. The filter sizes for the three convolutional layers were 4, 8, and 16. The kernel size was three and the activation function was the rectified linear unit (ReLU). A back propagation neural network (BPNN) model was created without helix matrix transformation and a convolution layer to demonstrate whether the helix matrix transformation combined with CNN algorithm works better. The areas under the receiver operating characteristic (ROC) curve of the CNN and BPNN models were 0.98 and 0.87, respectively. The accuracies of the CNN and BPNN models were 97.78 ± 0.08 and 86.50 ± 0.01, respectively, with a significant statistical difference (p < 0.001). The results suggested that helix matrix transformation combined with the CNN algorithm enabled the feature extraction of the bacterial MALDI-TOF MS spectrum, which might be a proposed solution to identify bacterial species.
Project description:Our aim was to evaluate the performance of matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), routinely used in the microbiology laboratory for bacterial identification, for bacterial typing in the setting of extended spectrum beta-lactamase producing Klebsiella pneumoniae (ESBL-KP) outbreak in the neonatal intensive care unit (NICU). Isolates from a 2011 outbreak in the NICU were retrieved from frozen stocks and analyzed by MALDI-TOF. The MALDI typing was compared with core genome multilocus sequence typing (cg-MLST). MALDI typing divided the 33 outbreak isolates into 2 clones: sequence type (ST)-290 and 405. These results were in complete agreement with cg-MLST results. The differentiation of the outbreak isolates into two clones correlated with the patients' location in the NICU, but also with their place of residence.Conclusion: Here, we show that MALDI-TOF MS, which has been integrated into the microbiology laboratory workflow for microbial species identification, can be secondarily used for epidemiological typing at no added cost.What is Known:• Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) is now routinely used in the microbiology laboratory for bacterial identificationWhat is New:• MALDI typing was used for outbreak investigation in the NICU and divided the outbreak isolates into two clones• MALDI-TOF MS may be secondarily used for epidemiological typing at no added cost.
Project description:Bacillus are aerobic spore-forming bacteria that are known to lead to specific diseases, such as anthrax and food poisoning. This study focuses on the characterization of these bacteria by the detection of lipids extracted from 33 well-characterized strains from the Bacillus and Brevibacillus genera, with the aim to discriminate between the different species. For the purpose of analysing the lipids extracted from these bacterial samples, two rapid physicochemical techniques were used: matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry (MALDI-TOF-MS) and liquid chromatography in conjunction with mass spectrometry (LC-MS). The findings of this investigation confirmed that MALDI-TOF-MS could be used to identify different bacterial lipids and, in combination with appropriate chemometrics, allowed for the discrimination between these different bacterial species, which was supported by LC-MS. The average correct classification rates for the seven species of bacteria were 62.23 and 77.03 % based on MALDI-TOF-MS and LC-MS data, respectively. The Procrustes distance for the two datasets was 0.0699, indicating that the results from the two techniques were very similar. In addition, we also compared these bacterial lipid MALDI-TOF-MS profiles to protein profiles also collected by MALDI-TOF-MS on the same bacteria (Procrustes distance, 0.1006). The level of discrimination between lipids and proteins was equivalent, and this further indicated the potential of MALDI-TOF-MS analysis as a rapid, robust and reliable method for the classification of bacteria based on different bacterial chemical components. Graphical abstract MALDI-MS has been successfully developed for the characterization of bacteria at the subspecies level using lipids and benchmarked against HPLC.
Project description:During the past 5 years, matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry (MS) has become a powerful tool for routine identification in many clinical laboratories. We analyzed our 11-year experience in routine identification of clinical isolates (40 months using MALDI-TOF MS and 91 months using conventional phenotypic identification [CPI]). Among the 286,842 clonal isolates, 284,899 isolates of 459 species were identified. The remaining 1,951 isolates were misidentified and required confirmation using a second phenotypic identification for 670 isolates and using a molecular technique for 1,273 isolates of 339 species. MALDI-TOF MS annually identified 112 species, i.e., 36 species/10,000 isolates, compared to 44 species, i.e., 19 species/10,000 isolates, for CPI. Only 50 isolates required second phenotypic identifications during the MALDI-TOF MS period (i.e., 4.5 reidentifications/10,000 isolates) compared with 620 isolates during the CPI period (i.e., 35.2/10,000 isolates). We identified 128 bacterial species rarely reported as human pathogens, including 48 using phenotypic techniques (22 using CPI and 37 using MALDI-TOF MS). Another 75 rare species were identified using molecular methods. MALDI-TOF MS reduced the time required for identification by 55-fold and 169-fold and the cost by 5-fold and 96-fold compared with CPI and gene sequencing, respectively. MALDI-TOF MS was a powerful tool not only for routine bacterial identification but also for identification of rare bacterial species implicated in human infectious diseases. The ability to rapidly identify bacterial species rarely described as pathogens in specific clinical specimens will help us to study the clinical burden resulting from the emergence of these species as human pathogens, and MALDI-TOF MS may be considered an alternative to molecular methods in clinical laboratories.
Project description:Matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) has gained popularity in recent years for rapid bacterial identification, mostly at the genus or species level. In this study, a rapid method to identify the Escherichia coli flagellar antigen (H antigen) at the subspecies level was developed using a MALDI-TOF MS platform with high specificity and sensitivity. Flagella were trapped on a filter membrane, and on-filter trypsin digestion was performed. The tryptic digests of each flagellin then were collected and analyzed by MALDI-TOF MS through peptide mass fingerprinting. Sixty-one reference strains containing all 53 H types and 85 clinical strains were tested and compared to serotyping designations. Whole-genome sequencing was used to resolve conflicting results between the two methods. It was found that DHB (2,5-dihydroxybenzoic acid) worked better than CHCA (?-cyano-4-hydroxycinnamic acid) as the matrix for MALDI-TOF MS, with higher confidence during protein identification. After method optimization, reference strains representing all 53 E. coli H types were identified correctly by MALDI-TOF MS. A custom E. coli flagellar/H antigen database was crucial for clearly identifying the E. coli H antigens. Of 85 clinical isolates tested by MALDI-TOF MS-H, 75 identified MS-H types (88.2%) matched results obtained from traditional serotyping. Among 10 isolates where the results of MALDI-TOF MS-H and serotyping did not agree, 60% of H types characterized by whole-genome sequencing agreed with those identified by MALDI-TOF MS-H, compared to only 20% by serotyping. This MALDI-TOF MS-H platform can be used for rapid and cost-effective E. coli H antigen identification, especially during E. coli outbreaks.
Project description:<h4>Aims</h4>The identification and differentiation of antibiotic-resistant bacteria by matrix-assisted laser desorption/ionization-time of flight-mass spectrometry (MALDI-TOF-MS) profiling remains a challenge due to the difficulty in detecting unique protein biomarkers associated with this trait. To expand the detectable proteome in antibiotic-resistant bacteria, we describe a method implementing offline LC protein separation/fractionation prior to MALDI-ToF-MS and top-down MALDI-ToF/ToF-MS (tandem MS or MS/MS) for the analysis of several antibiotic-resistant Escherichia coli isolates.<h4>Methods and results</h4>Coupling offline LC with MALDI-ToF-MS increased the number of detected protein signals in the typically analyzed mass regions (m/z 3000-20 000) by a factor of 13. Using the developed LC-MALDI-ToF-MS protocol in conjunction with supervised principal components analysis, we detected a protein biomarker at m/z 9355 which correlated to ?-lactam resistance among the E. coli bacteria tested. Implementing a top-down MALDI-ToF/ToF-MS approach, the prefractionated protein biomarker was inferred as a DNA-binding HU protein, likely translated from the bla<sub>CMY-2</sub> gene (encoding AmpC-type ?-lactamase) in the incompatibility plasmid complex A/C (IncA/C).<h4>Conclusions</h4>Our results demonstrate the utility of LC-MALDI-MS and MS/MS to extend the number of proteins detected and perform MALDI-accessible protein biomarker discovery in microorganisms.<h4>Significance and impact of the study</h4>This outcome is significant since it expands the detectable bacterial proteome via MALDI-ToF-MS.
Project description:Since the 1970s, the Planetary Protection Group at the Jet Propulsion Laboratory (JPL) has maintained an archive of spacecraft-associated bacterial isolates. Identification of these isolates was routinely performed by sequencing the 16S rRNA gene. Although this technique is an industry standard, it is time consuming and has poor resolving power for some closely related taxa. Matrix-assisted laser desorption/ionization time of flight (MALDI-TOF) mass spectrometry is widely used in clinical diagnostics and is a promising method to substitute standard 16S rRNA sequencing. However, manufacturer-provided databases lack the bacterial diversity found in spacecraft-assembly cleanrooms. This study reports the development of the first custom database of MALDI-TOF MS profiles of bacterial isolates obtained from spacecraft and associated cleanroom environments. With the use of this in-house database, 454 bacterial isolates were successfully identified in concurrence with their 16S rRNA sequence-based classifications. Additionally, MALDI-TOF MS resolved strain-level variations, identified potential novel species and distinguished between members of taxonomic groups, which is not possible using conventional 16S rRNA sequencing. MALDI-TOF MS has proved to be an accurate, high-throughput approach for real-time identification of bacterial isolates during the spacecraft assembly process.
Project description:Conventional methods are sometimes insufficient to identify human bacterial pathogens, and alternative techniques, often molecular, are required. Matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) identified with a valid score 45.9% of 410 clinical isolates from 207 different difficult-to-identify species having required 16S rRNA gene sequencing. MALDI-TOF MS might represent an alternative to 16S rRNA gene sequencing.
Project description:A method has been developed to identify the ?-subunit of Shiga toxin 2 (?-Stx2) from Escherichia coli O157:H7 using matrix-assisted laser desorption/ionization time-of-flight-time-of-flight tandem mass spectrometry (MALDI-TOF-TOF-MS/MS) and top-down proteomics using web-based software developed in-house. Expression of Stx2 was induced by culturing E. coli O157:H7 on solid agar supplemented with an antibiotic that elicits the bacterial SOS-response. Bacterial cell lysates were incubated in the presence of furin, a human enzyme, that cleaves ?-Stx2 into A1 (~28?kDa) and A2 (~5?kDa) protein fragments. A subsequent disulfide reduction step unlinked A1 from A2. MALDI-TOF-MS of the furin-digested/disulfide-reduced sample showed a peak at mass-to-charge (m/z) 5286 that corresponded to the A2 fragment. No peak was observed that corresponded to the A1 fragment although its presence was confirmed by bottom-up proteomics. The peak at m/z 5286 was definitively identified by MALDI-TOF-TOF-MS/MS and top-down proteomics as the A2 fragment of ?-Stx2.
Project description:In order to visualize the relationship between bacterial phylogeny and specialized metabolite production of bacterial colonies growing on nutrient agar, we developed IDBac-a low-cost and high-throughput matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) bioinformatics pipeline. IDBac software is designed for non-experts, is freely available, and capable of analyzing a few to thousands of bacterial colonies. Here, we present procedures for the preparation of bacterial colonies for MALDI-TOF MS analysis, MS instrument operation, and data processing and visualization in IDBac. In particular, we instruct users how to cluster bacteria into dendrograms based on protein MS fingerprints and interactively create Metabolite Association Networks (MANs) from specialized metabolite data.
Project description:Vibrio parahaemolyticus is a pathogenic marine bacterium that is the main causative agent of bacterial seafood-borne gastroenteritis in the United States. An increase in the frequency of V. parahaemolyticus-related infections during the last decade has been attributed to the emergence of an O3:K6 pandemic clone in 1995. The diversity of the O3:K6 pandemic clone and its serovariants has been examined using multiple molecular techniques including multilocus sequence analysis, pulsed-field gel electrophoresis, and group-specific PCR analysis. Matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) has become a powerful tool for rapidly distinguishing between related bacterial species. In the current study, we demonstrate the development of a whole-cell MALDI-TOF MS method for the distinction of V. parahaemolyticus from other Vibrio spp. We identified 30 peaks that were present only in the spectra of the V. parahaemolyticus strains examined in this study that may be developed as MALDI-TOF MS biomarkers for identification of V. parahaemolyticus. We detected variation in the MALDI-TOF spectra of V. parahaemolyticus strains isolated from different geographical locations and at different times. The MALDI-TOF MS spectra of the V. parahaemolyticus strains examined were distinct from those of the other Vibrio species examined including the closely related V. alginolyticus, V. harveyi, and V. campbellii. The results of this study demonstrate the first use of whole-cell MALDI-TOF MS analysis for the rapid identification of V. parahaemolyticus.