基于层级和全局特征结合的蛋白质序列EC编号预测

酶功能的识别对理解生命活动的机制、推进生命科学的发展有重要作用。然而现有的酶EC编号预测方法,并未充分利用蛋白质序列信息,在识别精度上仍有所不足。针对上述问题,本研究提出一种基于层级特征和全局特征的EC编号预测网络(EC number prediction network using hierarchical features and global features,ECPN-HFGF)。该方法首先通过残差网络提取蛋白质序列通用特征,并通过层级特征提取模块和全局特征提取模块进一步提取蛋白质序列的层级特征和全局特征,之后结合两种特征信息的预测结果,采用多任务学习框架,实现酶EC编号的精确预测。计算实验结果表明,ECPN-HFGF方法在蛋白质序列EC编号预测任务上性能最佳,宏观F1值和微观F1值分别达到95.5%和99.0%。ECPN-HFGF方法能有效结合蛋白质序列的层级特征和全局特征,快速准确预测蛋白质序列EC编号,比当前常用方法预测精确度更高,能够为酶学研究和酶工程应用的发展提供一种高效的思路和方法。

Significance of regenerating islet-derived type Ⅳ gene expression in gastroenterological cancers

The regenerating islet-derived members （Reg）, a group of small secretory proteins, which are involved in cell proliferation or differentiation in digestive organs, are upregulated in several gastrointestinal cancers, functioning as trophic or antiapoptotic factors. Regenerat- ing islet-derived type Ⅳ （RegⅣ）, a member of the Reg gene family, has been reported to be overexpressed in gastroenterological cancers. RegIV overexpression in tumor cells has been associated with carcinogen- esis, cell growth, survival and resistance to apoptosis. Cancer tissue expressing RegIV is generally associated with more malignant characteristics than that with- out such expression, and RegⅣ is considered a novel prognostic factor as well as diagnostic marker in some gastroenterological cancers. We previously investigated the expression levels of RegⅣ mRNA of 202 surgical colorectal cancer specimens with quantitative real-time reverse-transcriptase polymerase chain reaction and reported that a higher level of RegⅣ gene expression was a significant independent predictor of colorec- tal cancer. The biologic functions of RegⅣ protein in cancer tissue, associated with carcinogenesis, anti- apoptosis and invasiveness, are being elucidated by molecular investigations using transfection techniques or neutralizing antibodies of RegIV, and the feasibility of antibody therapy targeting RegIV is being assessed. These studies may lead to novel therapeutic strate- gies for gastroenterological cancers expressing RegⅣ. This review article summarizes the current information related to biological functions as well as clinical impor- tance of RegⅣ gene to clarify the significance of Reg~ expression in gastroenterological cancers.

Structure-based function prediction of the expanding mollusk tyrosinase family

Tyrosinase （Ty） is a common enzyme found in many different animal groups. In our previous study, genome sequencing revealed that the Ty family is expanded in the Pacific oyster （Crassostrea gigas）. Here, we examine the larger number of Ty family members in the Pacific oyster by high-level structure prediction to obtain more information about their function and evolution, especially the unknown role in biomineralization. We verified 12 Ty gene sequences from Crassostrea gigas genome and Pinctadafucata martensii transcriptome. By using phylogenetic analysis of these Tys with functionally known Tys from other molluscan species, eight subgroups were identified （CgTy_sl, CgTy s2, MolTy sl, MolTy-s2, MolTy-s3, PinTy-s 1, PinTy-s2 and PviTy）. Structural data and surface pockets of the dinuclear copper center in the eight subgroups of molluscan Ty were obtained using the latest versions of prediction online servers. Structural comparison with other Ty proteins from the protein databank revealed functionally important residues （HA1, HA2, HA3, HB1, HB2, HB3, Z l-Z9） and their location within these protein structures. The structural and chemical features of these pockets which may related to the substrate binding showed considerable variability among mollusks, which undoubtedly defines Ty substrate binding. Finally, we discuss the potential driving forces of Ty family evolution in mollusks. Based on these observations, we conclude that the Ty family has rapidly evolved as a consequence of substrate adaptation in mollusks.