Molecular Structure Databases Research Articles

Escherichia coli research on Raman measurement mechanism and diagnostic model

Escherichia coli (E. coli) is one of the most important pathogenic bacteria causing poultry diseases, characterized by a wide distribution range, rapid spread, and high mortality rate. Early diagnosis of E. coli in poultry feces provides the possibility for targeted treatment and rapid recovery of diseased poultry, and more importantly, prevents the rapid spread of pathogens among densely bred poultry. In order to implement rapid, low-cost, and high-frequency detection of E. coli, this study explored the feasibility of Raman spectroscopy. Firstly, theoretical configurations and density functional calculations of N-acetylmuramic acid and N-acetylglucosamine in the cell wall of E. coli were performed. Then, Raman measurement models for E. coli were established based on two feature extraction methods (Successive Projections Algorithm, Competitive Adaptive Reweighted Sampling) and four modeling methods (Random Forest Algorithm, Convolutional Neural Networks, Back Propagation Neural Networks, Radial Basis Function). Finally, a method based on the extraction of Raman spectral features using density functional theory was determined to optimize the existing models, and it was demonstrated that this feature variable extraction method improved the accuracy of all four measurement models to some extent. Ultimately, the optimal model, the improved SPA-RF, was obtained through comparative analysis, with an accuracy, precision, recall, specificity, FNR, FDR, and AUC of 98.38%, 98.61%, 99.83%, 88.08%, 0.81%, 11.82%, and 1, respectively. This study reports an early method for the early treatment of E. coli diseases and provides a molecular structure database for studying N-acetylmuramic acid and N-acetylglucosamine, as well as a basis for vibrational spectroscopy detection of E. coli diseases, promoting the application of Raman spectroscopy technology in the diagnosis of livestock diseases.

Mining Secondary Metabolites of Chassalia curviflora Leaves Using SIRIUS and High-Resolution Mass Spectrometry Data

Chassalia curviflora (Wall ex Kurz.) Thwaites is a flowering plant from Rubiaceae family. Previous phytochemical investigation on Chassalia species led to the isolation of benzoic acid, benzoquinone derivatives, flavonoids, monoterpenoid indole alkaloids and its essential oil composition. However, these are still scarce in terms of its scientific report on studied species. This study was carried out to investigate fast screening of secondary metabolites by using data-dependent acquisition (DDA) LC-MS/MS analysis of methanolic extract of the leaf of the C. curviflora in positive ionization mode. The data were analysed using computational mass spectrometry method from molecular structure database search implemented in Sirius. This led in discovering of 13 known secondary metabolites from a distinct tandem MS data. De novo molecular formula annotations and predictions from SIRIUS suggested five major class compounds classified as coumarin, benzofuran, fatty acids, esters and triterpenes. Stigmasterol 12 and β-sitosterol 13 were successfully isolated, thus used to verify the data obtained from the Sirius. With our results, we advocate the MS-based approach as a useful starting method for the dereplication of compounds from the complex crude extract of plants.

MAD HATTER Correctly Annotates 98% of Small Molecule Tandem Mass Spectra Searching in PubChem.

Metabolites provide a direct functional signature of cellular state. Untargeted metabolomics usually relies on mass spectrometry, a technology capable of detecting thousands of compounds in a biological sample. Metabolite annotation is executed using tandem mass spectrometry. Spectral library search is far from comprehensive, and numerous compounds remain unannotated. So-called in silico methods allow us to overcome the restrictions of spectral libraries, by searching in much larger molecular structure databases. Yet, after more than a decade of method development, in silico methods still do not reach the correct annotation rates that users would wish for. Here, we present a novel computational method called Mad Hatter for this task. Mad Hatter combines CSI:FingerID results with information from the searched structure database via a metascore. Compound information includes the melting point, and the number of words in the compound description starting with the letter 'u'. We then show that Mad Hatter reaches a stunning 97.6% correct annotations when searching PubChem, one of the largest and most comprehensive molecular structure databases. Unfortunately, Mad Hatter is not a real method. Rather, we developed Mad Hatter solely for the purpose of demonstrating common issues in computational method development and evaluation. We explain what evaluation glitches were necessary for Mad Hatter to reach this annotation level, what is wrong with similar metascores in general, and why metascores may screw up not only method evaluations but also the analysis of biological experiments. This paper may serve as an example of problems in the development and evaluation of machine learning models for metabolite annotation.

Deep kernel learning improves molecular fingerprint prediction from tandem mass spectra.

MotivationUntargeted metabolomics experiments rely on spectral libraries for structure annotation, but these libraries are vastly incomplete; in silico methods search in structure databases, allowing us to overcome this limitation. The best-performing in silico methods use machine learning to predict a molecular fingerprint from tandem mass spectra, then use the predicted fingerprint to search in a molecular structure database. Predicted molecular fingerprints are also of great interest for compound class annotation, de novo structure elucidation, and other tasks. So far, kernel support vector machines are the best tool for fingerprint prediction. However, they cannot be trained on all publicly available reference spectra because their training time scales cubically with the number of training data.ResultsWe use the Nyström approximation to transform the kernel into a linear feature map. We evaluate two methods that use this feature map as input: a linear support vector machine and a deep neural network (DNN). For evaluation, we use a cross-validated dataset of 156 017 compounds and three independent datasets with 1734 compounds. We show that the combination of kernel method and DNN outperforms the kernel support vector machine, which is the current gold standard, as well as a DNN on tandem mass spectra on all evaluation datasets.Availability and implementationThe deep kernel learning method for fingerprint prediction is part of the SIRIUS software, available at https://bio.informatik.uni-jena.de/software/sirius.

Predicting a Molecular Fingerprint from an Electron Ionization Mass Spectrum with Deep Neural Networks.

Electron ionization-mass spectrometry (EI-MS) hyphenated to gas chromatography (GC) is the workhorse for analyzing volatile compounds in complex samples. The spectral matching method can only identify compounds within the spectral database. In response, we present a deep-learning-based approach (DeepEI) for structure elucidation of an unknown compound with its EI-MS spectrum. DeepEI employs deep neural networks to predict molecular fingerprints from an EI-MS spectrum and searches the molecular structure database with the predicted fingerprints. We evaluated DeepEI with MassBank spectra, and the results indicate DeepEI is an effective identification method. In addition, DeepEI can work cooperatively with database spectral matching and NEIMS (fingerprint to spectrum method) to improve identification accuracy.

Accurate and Efficient Determination of Unknown Metabolites in Metabolomics by NMR-Based Molecular Motif Identification.

Knowledge of the chemical identity of metabolite molecules is critical for the understanding of the complex biological systems to which they belong. Since metabolite identities and their concentrations are often directly linked to the phenotype, such information can be used to map biochemical pathways and understand their role in health and disease. A very large number of metabolites however are still unknown; i.e., their spectroscopic signatures do not match those in existing databases, suggesting unknown molecule identification is both imperative and challenging. Although metabolites are structurally highly diverse, the majority shares a rather limited number of structural motifs, which are defined by sets of 1H and 13C chemical shifts of the same spin system. This allows one to characterize unknown metabolites by a divide-and-conquer strategy that identifies their structural motifs first. Here, we present the structural motif-based approach "SUMMIT Motif" for the de novo identification of unknown molecular structures in complex mixtures, without the need for extensive purification, using NMR in tandem with two newly curated NMR molecular structural motif metabolomics databases (MSMMDBs). For the identification of structural motif(s), first, the 1H and 13C chemical shifts of all the individual spin systems are extracted from 2D and 3D NMR spectra of the complex mixture. Next, the molecular structural motifs are identified by querying these chemical shifts against the new MSMMDBs. One database, COLMAR MSMMDB, was derived from experimental NMR chemical shifts of known metabolites taken from the COLMAR metabolomics database, while the other MSMMDB, pNMR MSMMDB, is based on predicted chemical shifts of metabolites of several existing large metabolomics databases. For molecules consisting of multiple spin systems, spin systems are connected via long-range scalar J-couplings. When this motif-based identification method was applied to the hydrophilic extract of mouse bile fluid, two unknown metabolites could be successfully identified. This approach is both accurate and efficient for the identification of unknown metabolites and hence enables the discovery of new biochemical processes and potential biomarkers.

Identification of metabolites from tandem mass spectra with a machine learning approach utilizing structural features.

MotivationUntargeted mass spectrometry (MS/MS) is a powerful method for detecting metabolites in biological samples. However, fast and accurate identification of the metabolites’ structures from MS/MS spectra is still a great challenge.ResultsWe present a new analysis method, called SubFragment-Matching (SF-Matching) that is based on the hypothesis that molecules with similar structural features will exhibit similar fragmentation patterns. We combine information on fragmentation patterns of molecules with shared substructures and then use random forest models to predict whether a given structure can yield a certain fragmentation pattern. These models can then be used to score candidate molecules for a given mass spectrum. For rapid identification, we pre-compute such scores for common biological molecular structure databases. Using benchmarking datasets, we find that our method has similar performance to CSI: FingerID and those very high accuracies can be achieved by combining our method with CSI: FingerID. Rarefaction analysis of the training dataset shows that the performance of our method will increase as more experimental data become available.Availability and implementationSF-Matching is available from http://www.bork.embl.de/Docu/sf_matching.Supplementary information Supplementary data are available at Bioinformatics online.

SIRIUS 4: a rapid tool for turning tandem mass spectra into metabolite structure information.

Mass spectrometry is a predominant experimental technique in metabolomics and related fields, but metabolite structural elucidation remains highly challenging. We report SIRIUS 4 (https://bio.informatik.uni-jena.de/sirius/), which provides a fast computational approach for molecular structure identification. SIRIUS 4 integrates CSI:FingerID for searching in molecular structure databases. Using SIRIUS 4, we achieved identification rates of more than 70% on challenging metabolomics datasets.

Generation of Pairwise Potentials Using Multidimensional Data Mining.

The rapid development of molecular structural databases provides the chemistry community access to an enormous array of experimental data that can be used to build and validate computational models. Using radial distribution functions collected from experimentally available X-ray and NMR structures, a number of so-called statistical potentials have been developed over the years using the structural data mining strategy. These potentials have been developed within the context of the two-particle Kirkwood equation by extending its original use for isotropic monatomic systems to anisotropic biomolecular systems. However, the accuracy and the unclear physical meaning of statistical potentials have long formed the central arguments against such methods. In this work, we present a new approach to generate molecular energy functions using structural data mining. Instead of employing the Kirkwood equation and introducing the "reference state" approximation, we model the multidimensional probability distributions of the molecular system using graphical models and generate the target pairwise Boltzmann probabilities using the Bayesian field theory. Different from the current statistical potentials that mimic the "knowledge-based" PMF based on the 2-particle Kirkwood equation, the graphical-model-based structure-derived potential developed in this study focuses on the generation of lower-dimensional Boltzmann distributions of atoms through reduction of dimensionality. We have named this new scoring function GARF, and in this work we focus on the mathematical derivation of our novel approach followed by validation studies on its ability to predict protein-ligand interactions.

Bayesian networks for mass spectrometric metabolite identification via molecular fingerprints.

MotivationMetabolites, small molecules that are involved in cellular reactions, provide a direct functional signature of cellular state. Untargeted metabolomics experiments usually rely on tandem mass spectrometry to identify the thousands of compounds in a biological sample. Recently, we presented CSI:FingerID for searching in molecular structure databases using tandem mass spectrometry data. CSI:FingerID predicts a molecular fingerprint that encodes the structure of the query compound, then uses this to search a molecular structure database such as PubChem. Scoring of the predicted query fingerprint and deterministic target fingerprints is carried out assuming independence between the molecular properties constituting the fingerprint.ResultsWe present a scoring that takes into account dependencies between molecular properties. As before, we predict posterior probabilities of molecular properties using machine learning. Dependencies between molecular properties are modeled as a Bayesian tree network; the tree structure is estimated on the fly from the instance data. For each edge, we also estimate the expected covariance between the two random variables. For fixed marginal probabilities, we then estimate conditional probabilities using the known covariance. Now, the corrected posterior probability of each candidate can be computed, and candidates are ranked by this score. Modeling dependencies improves identification rates of CSI:FingerID by 2.85 percentage points.Availability and implementationThe new scoring Bayesian (fixed tree) is integrated into SIRIUS 4.0 (https://bio.informatik.uni-jena.de/software/sirius/).

MetExpert: An expert system to enhance gas chromatography‒mass spectrometry-based metabolite identifications

Gas chromatography‒mass spectrometry (GCMS) is an important analytical technique in metabolomics studies and has been routinely used for metabolite profiling of biological samples. Spectral matching to databases of authentic compounds are the preferred tools for metabolite identifications; however, the metabolic coverage of these databases is still limited compared to the number of known metabolites. Several computational tools have been developed to facilitate the interpretation of MS data but unfortunately most of them have limited applicability to GCMS-based metabolite identification. In this paper, we introduce a computer-aided, metabolite expert system called MetExpert which emulates the metabolite-identification ability of a human expert using orthogonal datasets including molecular formulas, retention indices, and EI-MS spectra to characterize the molecular structures. This system integrates four modules including in silico derivatization, metabolite-likeness evaluation, retention prediction, and substructure prediction. In silico derivatization increases the searchable chemical space for TMS-derivatized metabolites many of which are absent in molecular structure databases. Metabolite-likeness evaluations are an efficient approach to select metabolite-like molecules when querying large databases such as PubChem. An artificial neutral network then establishes the quantitative structure‒retention relationships for the accurate prediction of RIs that further refines the candidate molecules. In addition, PLS-DA models establish quantitative structure‒spectra relationships for the prediction of metabolite substructures. Finally, weighted scoring of three orthogonal evaluations increases the identification rates. MetExpert outperformed current state-of-the-art methods such as MetFrag and CFM-ID for ranking the correct identifications. While spectral comparisons with chemical standards or de novo structural elucidations are necessary to validate the predictions, MetExpert provides an efficient and effective approach to prioritize the candidates.

Pharmacophore-based discovery of 2-(phenylamino)aceto-hydrazides as potent eosinophil peroxidase (EPO) inhibitors

There is an increasing interest in developing novel eosinophil peroxidase (EPO) inhibitors, in order to provide new treatment strategies against chronic inflammatory and neurodegenerative diseases caused by eosinophilic disorder. Within this study, a ligand-based pharmacophore model for EPO inhibitors was generated and used for in silico screening of large 3 D molecular structure databases, containing more than 4 million compounds. Hits obtained were clustered and a total of 277 compounds were selected for biological assessment. A class of 2-(phenyl)amino-aceto-hydrazides with different substitution pattern on the aromatic ring was found to contain the most potent EPO inhibitors, exhibiting IC50 values down to 10 nM. The generated pharmacophore model therefore, represents a valuable tool for the selection of compounds for biological testing. The compounds identified as potent EPO inhibitors will serve to initiate a hit to lead and lead optimisation program for the development of new therapeutics against eosinophilic disorders.

Practical application of in silico fragmentation based residue screening with ion mobility high-resolution mass spectrometry.

A screening concept for residues in complex matrices based on liquid chromatography coupled to ion mobility high-resolution mass spectrometry LC/IMS-HRMS is presented. The comprehensive four-dimensional data (chromatographic retention time, drift time, mass-to-charge and ion abundance) obtained in data-independent acquisition (DIA) mode was used for data mining. An in silico fragmenter utilizing a molecular structure database was used for suspect screening, instead of targeted screening with reference substances. The utilized data-independent acquisition mode relies on the MSE concept; where two constantly alternating HRMS scans (low and high fragmentation energy) are acquired. Peak deconvolution and drift time alignment of ions from the low (precursor ion) and high (product ion) energy scan result in relatively clean product ion spectra. A bond dissociation in silico fragmenter (MassFragment) supplied with mol files of compounds of interest was used to explain the observed product ions of each extracted candidate component (chromatographic peak). Two complex matrices (fish and bovine liver extract) were fortified with 98 veterinary drugs. Out of 98 screened compounds 94 could be detected with the in silico based screening approach. The high correlation among drift time and m/z value of equally charged ions was utilized for an orthogonal filtration (ranking). Such an orthogonal ion mobility based filter removes multiply charged ions (e.g. peptides and proteins from the matrix) as well as noise and artefacts. Most significantly, this filtration dramatically reduces false positive findings but hardly increases false negative findings. The proposed screening approach may offer new possibilities for applications where reference compounds are hardly or not at all commercially available. Such areas may be the analysis of metabolites of drugs, pyrrolizidine alkaloids, marine toxins, derivatives of sildenafil or novel designer drugs (new psychoactive substances). Copyright © 2017 John Wiley & Sons, Ltd.

Enhanced metabolite annotation via dynamic retention time prediction: Steroidogenesis alterations as a case study.

The development of metabolomics based on ultra-high pressure liquid chromatography coupled to high-resolution mass spectrometry (UHPLC-HRMS) now allows hundreds to thousands of metabolites to be simultaneously monitored in biological matrices. In that context, bioinformatics and multivariate data analysis (MVA) play a crucial role in the detection of relevant alteration patterns. However, sound biological interpretations must necessarily be supported by metabolite identifications to be definitive or at least have a high degree of confidence. Each compound, should be characterised by unique molecular properties. Among them, the exact mass and the chromatographic retention time are recognised as major and complementary criteria for compound identification. While the former is easily derived from the molecular structure, building generic and accurate retention time open databases still constitutes a critical issue because of the vast diversity of instruments, stationary phases and operating conditions in UHPLC-HRMS. Because several hits matching a molecular formula obtained from an exact mass and an isotopic pattern are often generated for each analyte, this methodology rarely allows a unique and unambiguous molecular identity to be gained. This work aims to provide a flexible solution to facilitate reliable compound annotation based on retention time in reversed-phase liquid chromatography (RPLC). It proposes an innovative approach based on the chromatographic linear solvent strength (LSS) theory, allowing retention times under any gradient conditions at fixed temperature, stationary phase and mobile phase type to be predicted. Starting from a subset of the Human Metabolite Database (HMDB), a new dynamic database involving LSS parameters was developed. A real case study involving steroidogenesis alterations due to forskolin exposure was conducted using the adrenal H295R OECD reference cell model for endocrine disruptor screening. The prediction of retention times was successfully achieved, facilitating steroid identification. An automated procedure which implements the compound annotation levels encouraged by the Metabolite Standard Initiative (MSI) and the Coordination of Standards in Metabolomics (COSMOS) was also developed to speed up the process and enhance the data reusability.

Homology modeling and virtual screening studies of FGF-7 protein-a structure-based approach to design new molecules against tumor angiogenesis.

Keratinocyte growth factor (KGF) protein is a member of the fibroblast growth factor (FGF) family, which is also known as FGF-7. The FGF-7 plays an important role in tumor angiogenesis. In the present work, FGF-7 is treated as a potential therapeutic target to prevent angiogenesis in cancerous tissue. Computational techniques are applied to evaluate and validate the 3D structure of FGF-7 protein. The active site region of the FGF-7 protein is identified based on hydrophobicity calculations using CASTp and Q-site Finder active site prediction tools. The protein-protein docking study of FGF-7 with its natural receptor FGFR2b is carried out to confirm the active site region in FGF-7. The amino acid residues Asp34, Arg67, Glu116, and Thr194 in FGF-7 interact with the receptor protein (FGFR2b). A grid is generated at the active site region of FGF-7 using Glide module of Schrödinger suite. Subsequently, a virtual screening study is carried out at the active site using small molecular structural databases to identify the ligand molecules. The binding interactions of the ligand molecules, with piperazine moiety as a pharmacophore, are observed at Arg67 and Glu149 residues of the FGF-7 protein. The identified ligand molecules against the FGF-7 protein show permissible pharmacokinetic properties (ADME). The ligand molecules with good docking scores and satisfactory pharmacokinetic properties are prioritized and identified as novel ligands for the FGF-7 protein in cancer therapy.

Fast metabolite identification with Input Output Kernel Regression.

Motivation: An important problematic of metabolomics is to identify metabolites using tandem mass spectrometry data. Machine learning methods have been proposed recently to solve this problem by predicting molecular fingerprint vectors and matching these fingerprints against existing molecular structure databases. In this work we propose to address the metabolite identification problem using a structured output prediction approach. This type of approach is not limited to vector output space and can handle structured output space such as the molecule space.Results: We use the Input Output Kernel Regression method to learn the mapping between tandem mass spectra and molecular structures. The principle of this method is to encode the similarities in the input (spectra) space and the similarities in the output (molecule) space using two kernel functions. This method approximates the spectra-molecule mapping in two phases. The first phase corresponds to a regression problem from the input space to the feature space associated to the output kernel. The second phase is a preimage problem, consisting in mapping back the predicted output feature vectors to the molecule space. We show that our approach achieves state-of-the-art accuracy in metabolite identification. Moreover, our method has the advantage of decreasing the running times for the training step and the test step by several orders of magnitude over the preceding methods.Availability and implementation:Contact: celine.brouard@aalto.fiSupplementary information: Supplementary data are available at Bioinformatics online.

Mining molecular structure databases: Identification of small molecules based on fragmentation mass spectrometry data.

Mass spectrometry (MS) is a key technology for the analysis of small molecules. For the identification and structural elucidation of novel molecules, new approaches beyond straightforward spectral comparison are required. In this review, we will cover computational methods that help with the identification of small molecules by analyzing fragmentation MS data. We focus on the four main approaches to mine a database of metabolite structures, that is rule-based fragmentation spectrum prediction, combinatorial fragmentation, competitive fragmentation modeling, and molecular fingerprint prediction. © 2016 Wiley Periodicals, Inc. Mass Spec Rev 36:624-633, 2017.

The development of an integrated system for structural chemical information

Molecular and reaction structural databases on chemistry are considered. The history of how the software-technological complex of the Database of Structural Information on Chemistry of the VINITI RAS developed is described. The potential of searching for structural information based on the generalization of structural diagrams is shown.

Searching molecular structure databases with tandem mass spectra using CSI:FingerID

Metabolites provide a direct functional signature of cellular state. Untargeted metabolomics experiments usually rely on tandem MS to identify the thousands of compounds in a biological sample. Today, the vast majority of metabolites remain unknown. We present a method for searching molecular structure databases using tandem MS data of small molecules. Our method computes a fragmentation tree that best explains the fragmentation spectrum of an unknown molecule. We use the fragmentation tree to predict the molecular structure fingerprint of the unknown compound using machine learning. This fingerprint is then used to search a molecular structure database such as PubChem. Our method is shown to improve on the competing methods for computational metabolite identification by a considerable margin.

Protein structure reports in JIMD ‐ an important enhancement of journal scope

Protein structure reports in JIMD ‐ an important enhancement of journal scope