Complete genome analysis of pathogenic Metschnikowia bicuspidata strain MQ2101 isolated from diseased ridgetail white prawn, Exopalaemon carinicauda.

BACKGROUND: Metschnikowia bicuspidata is a pathogenic yesst that can cause disease in many different economic aquatic animal species. In recent years, there was a new disease outbreak in ridgetail white prawn (Exopalaemon carinicauda) in coastal areas of Jiangsu Province China that was referred to as zombie disease by local farmers. The pathogen was first isolated and identified as M. bicuspidata. Although the pathogenicity and pathogenesis of this pathogen in other animals have been reported in some previous studies, research on its molecular mechanisms is still very limited. Therefore, a genome-wide study is necessary to better understand the physiological and pathogenic mechanisms of M. bicuspidata. RESULT: In this study, we obtained a pathogenic strain, MQ2101, of M. bicuspidata from diseased E. carinicauda and sequenced its whole genome. The size of the whole genome was 15.98 Mb, and it was assembled into 5 scaffolds. The genome contained 3934 coding genes, among which 3899 genes with biological functions were annotated in multiple underlying databases. In KOG database, 2627 genes were annotated, which were categorized into 25 classes including general function prediction only, posttranslational modification, protein turnover, chaperones, and signal transduction mechanisms. In KEGG database, 2493 genes were annotated, which were categorized into five classes, including cellular processes, environmental information processing, genetic information processing, metabolism and organismal systems. In GO database, 2893 genes were annotated, which were mainly classified in cell, cell part, cellular processes and metabolic processes. There were 1055 genes annotated in the PHI database, accounting for 26.81% of the total genome, among which 5 genes were directly related to pathogenicity (identity ≥ 50%), including hsp90, PacC, and PHO84. There were also some genes related to the activity of the yeast itself that could be targeted by antiyeast drugs. Analysis based on the DFVF database showed that strain MQ2101 contained 235 potential virulence genes. BLAST searches in the CAZy database showed that strain MQ2101 may have a more complex carbohydrate metabolism system than other yeasts of the same family. In addition, two gene clusters and 168 putative secretory proteins were predicted in strain MQ2101, and functional analysis showed that some of the secretory proteins may be directly involved in the pathogenesis of the strain. Gene family analysis with five other yeasts revealed that strain MQ2101 has 245 unique gene families, including 274 genes involved in pathogenicity that could serve as potential targets. CONCLUSION: Genome-wide analysis elucidated the pathogenicity-associated genes of M. bicuspidate while also revealing a complex metabolic mechanism and providing putative targets of action for the development of antiyeast drugs for this pathogen. The obtained whole-genome sequencing data provide an important theoretical basis for transcriptomic, proteomic and metabolic studies of M. bicuspidata and lay a foundation for defining its specific mechanism of host infestation.

[Classification technique for hyperspectral image based on subspace of bands feature extraction and LS-SVM].

The present paper proposes a novel hyperspectral image classification algorithm based on LS-SVM (least squares support vector machine). The LS-SVM uses the features extracted from subspace of bands (SOB). The maximum noise fraction (MNF) method is adopted as the feature extraction method. The spectral correlations of the hyperspectral image are used in order to divide the feature space into several SOBs. Then the MNF is used to extract characteristic features of the SOBs. The extracted features are combined into the feature vector for classification. So the strong bands correlation is avoided and the spectral redundancies are reduced. The LS-SVM classifier is adopted, which replaces inequality constraints in SVM by equality constraints. So the computation consumption is reduced and the learning performance is improved. The proposed method optimizes spectral information by feature extraction and reduces the spectral noise. The classifier performance is improved. Experimental results show the superiorities of the proposed algorithm.