RESUMEN
Broadening coverage in fatty acid (FA) analysis benefits the understanding of metabolic regulation in biological system. However, the limited access of chemical standards makes it challenging. In this work, we introduced a simulation assisted strategy to analyze short-, medium-, long- and very-long-chain fatty acids beyond the use of chemical standards. This targeted analysis in selected reaction monitoring (SRM) mode incorporated 3-nitrophenylhydrazine derivatization and mathematical simulation of ion transitions, collision energies, RF values and retention times to identify and quantify the fatty acids without chemical standards. Serum analysis using high resolution mass spectrometry coupled with paired labeling was employed to refine the computational retention times. Based on the simulation, 116 free fatty acids from C1 to C24 were covered in a single analysis on use of 34 standard chemicals. Background interference is commonly observed in fatty acid analysis. For certain fatty acids, e.g. acetic acid or palmitic acid, reliable quantitation is largely restricted by contamination level instead of detection limit. Therefore, the background interference and quantifiable serum volume required for each fatty acid were also evaluated. At least 20 µL serum was suggested to cover most molecules. Using this approach, a total of 66 free fatty acids with various chain lengths and saturations were detected in NTCP knockout mice serum, of which 34 FAs were confirmed by chemical standards and 32 FAs were potentially assigned based on the simulation. Gender dependent fatty acid regulation was observed by NTCP knockout. This work provides a unique strategy that enables to broaden the fatty acid coverage with the absence of chemical standards and is applicable to other derivatizations.
Asunto(s)
Cromatografía Liquida/métodos , Ácidos Grasos no Esterificados/sangre , Ácidos Grasos no Esterificados/química , Espectrometría de Masas en Tándem/métodos , Animales , Simulación por Computador , Análisis Discriminante , Femenino , Modelos Lineales , Masculino , Ratones , Ratones Noqueados , Reproducibilidad de los Resultados , Sensibilidad y EspecificidadRESUMEN
Single-particle electron cryomicroscopy (cryo-EM) involves estimating a set of parameters for each particle image and reconstructing a 3D density map; robust algorithms with accurate parameter estimation are essential for high resolution and automation. We introduce a particle-filter algorithm for cryo-EM, which provides high-dimensional parameter estimation through a posterior probability density function (PDF) of the parameters given in the model and the experimental image. The framework uses a set of random support points to represent such a PDF and assigns weighting coefficients not only among the parameters of each particle but also among different particles. We implemented the algorithm in a new program named THUNDER, which features self-adaptive parameter adjustment, tolerance to bad particles, and per-particle defocus refinement. We tested the algorithm by using cryo-EM datasets for the cyclic-nucleotide-gated (CNG) channel, the proteasome, ß-galactosidase, and an influenza hemagglutinin (HA) trimer, and observed substantial improvement in resolution.
Asunto(s)
Algoritmos , Microscopía por Crioelectrón/métodos , Procesamiento de Imagen Asistido por Computador/métodos , Imagenología Tridimensional/métodos , Programas Informáticos , Canales Catiónicos Regulados por Nucleótidos Cíclicos/ultraestructura , Glicoproteínas Hemaglutininas del Virus de la Influenza/ultraestructura , Humanos , Complejo de la Endopetidasa Proteasomal/ultraestructura , beta-Galactosidasa/ultraestructuraRESUMEN
BACKGROUND: Tuberculous pleural effusion (TPE) and malignant pleural effusion (MPE) are the 2 most frequent causes of exudative pleural effusions (PEs). However, the clinical differentiation is challenging. METHODS: Metabolic signatures in pleural effusion from 156 patients were profiled. An integrated semi-targeted metabolomics platform was incorporated for high throughput metabolite identification and quantitation. In this platform, orbitrap based mass spectrometry with data dependent MS/MS acquisition was applied in the analysis. In-house database containing ~1000MS/MS spectra were established and "MetaInt" was developed for metabolite alignment. RESULTS: Using this strategy, lower levels of amino acids, citric acid cycle intermediates and free fatty acids accompanied with elevated acyl-carnitines and bile acids were observed, demonstrating increased energy expenditure caused by TPE. Kynurenine pathway from tryptophan was significantly enhanced in TPE. The ratio of tryptophan/kynurenine exhibited decent performance in differentiating TPE from MPE with sensitivity of 92.7% and specificity of 86.1%. After two further independent validations, it turns out that the ratio of tryptophan/kynurenine can be applied confidently as a potential biomarker together with adenosine deaminase (ADA) for clinical diagnosis of TPE. CONCLUSIONS: Conclusively, the integrated in-house platform for high throughput semi-targeted metabolomics analysis reliably identified great potential of tryptophan/kynurenine ratio as a novel diagnostic biomarker to distinguish pleural effusion caused by tuberculosis and malignancy.
Asunto(s)
Metabolómica , Derrame Pleural Maligno/metabolismo , Tuberculosis/metabolismo , Biomarcadores/análisis , Humanos , Espectrometría de Masas , Análisis Multivariante , Derrame Pleural Maligno/diagnóstico , Curva ROC , Tuberculosis/diagnósticoRESUMEN
Coping with scarcity of labeled data is a common problem in sound classification tasks. Approaches for classifying sounds are commonly based on supervised learning algorithms, which require labeled data which is often scarce and leads to models that do not generalize well. In this paper, we make an efficient combination of confidence-based Active Learning and Self-Training with the aim of minimizing the need for human annotation for sound classification model training. The proposed method pre-processes the instances that are ready for labeling by calculating their classifier confidence scores, and then delivers the candidates with lower scores to human annotators, and those with high scores are automatically labeled by the machine. We demonstrate the feasibility and efficacy of this method in two practical scenarios: pool-based and stream-based processing. Extensive experimental results indicate that our approach requires significantly less labeled instances to reach the same performance in both scenarios compared to Passive Learning, Active Learning and Self-Training. A reduction of 52.2% in human labeled instances is achieved in both of the pool-based and stream-based scenarios on a sound classification task considering 16,930 sound instances.