Mucin-type O-glycosylation is one of the most common post-translational modifications of proteins. This glycosylation is initiated in the Golgi by the addition of the sugar N-acetylgalactosamine (GalNAc) onto protein Ser and Thr residues by a family of polypeptide GalNAc transferases. In humans, there are 20 isoforms that are differentially expressed across tissues that serve multiple important biological roles. Using random peptide substrates, isoform specific amino acid preferences have been obtained in the form of enhancement values (EV). These EVs alone have previously been used to predict O-glycosylation sites via the web based ISOGlyP (Isoform Specific O-Glycosylation Prediction) tool. Here, we explore additional protein features to determine whether these can complement the random peptide derived enhancement values and increase the predictive power of ISOGlyP. The inclusion of additional protein substrate features (such as secondary structure and surface accessibility) was found to increase sensitivity with minimal loss of specificity, when tested with three different published in vivo O-glycoproteomics data sets, thus increasing the overall accuracy of the ISOGlyP predictions.

中文翻译：

---

使用衍生蛋白质特征的增强值产品预测粘蛋白型 O-糖基化

粘蛋白型O-糖基化是蛋白质最常见的翻译后修饰之一。这种糖基化在高尔基体中通过多肽 GalNAc 转移酶家族将糖 N-乙酰半乳糖胺 (GalNAc) 添加到蛋白质 Ser 和 Thr 残基上而启动。在人类中，有 20 种异构体在组织中差异表达，这些异构体具有多种重要的生物学作用。使用随机肽底物，以增强值 (EV) 的形式获得了异构体特定的氨基酸偏好。这些 EV 之前已被用于通过基于 Web 的 ISOGlyP（异构体特异性 O-糖基化预测）工具预测 O-糖基化位点。这里，我们探索其他蛋白质特征以确定这些是否可以补充随机肽衍生的增强值并增加 ISOGlyP 的预测能力。当使用三个不同的已发表的体内 O-糖蛋白质组学数据集进行测试时，发现包含额外的蛋白质底物特征（例如二级结构和表面可及性）可以提高灵敏度，同时将特异性损失降至最低，从而提高 ISOGlyP 预测的整体准确性.