School of Engineering and Physics, The University of the South Pacific, Suva, Fiji.
School of Electrical and Electronics Engineering, Fiji National University, Suva, Fiji.
BMC Bioinformatics. 2019 Feb 4;19(Suppl 13):378. doi: 10.1186/s12859-018-2396-7.
Molecular Recognition Features (MoRFs) are short protein regions present in intrinsically disordered protein (IDPs) sequences. MoRFs interact with structured partner protein and upon interaction, they undergo a disorder-to-order transition to perform various biological functions. Analyses of MoRFs are important towards understanding their function.
Performance is reported using the MoRF dataset that has been previously used to compare the other existing MoRF predictors. The performance obtained in this study is equivalent to the benchmarked OPAL predictor, i.e., OPAL achieved AUC of 0.815, whereas the model in this study achieved AUC of 0.819 using TEST set.
Achieving comparable performance, the proposed method can be used as an alternative approach for MoRF prediction.
分子识别特征（MoRFs）是存在于无规卷曲蛋白质（IDPs）序列中的短蛋白质区域。MoRFs 与结构上的伙伴蛋白相互作用，在相互作用时，它们经历从无序到有序的转变，以执行各种生物功能。对 MoRFs 的分析对于理解它们的功能很重要。
使用先前用于比较其他现有 MoRF 预测器的 MoRF 数据集报告性能。本研究中获得的性能与基准 OPAL 预测器相当，即 OPAL 获得的 AUC 为 0.815，而本研究中的模型在使用 TEST 集时获得的 AUC 为 0.819。
在达到可比性能的情况下，所提出的方法可以用作 MoRF 预测的替代方法。
