Department of Computer Science, College of Engineering, City University of Hong Kong, Kowloon Tong, Hong Kong.
Methods Mol Biol. 2021;2212:307-323. doi: 10.1007/978-1-0716-0947-7_19.
Epistasis is a challenge in prediction, classification, and suspicion of human genetic diseases. Many technologies, methods, and tools have been developed for epistasis detection. Multifactor dimensionality reduction (MDR) is the method commonly used in epistasis detection. It uses two class groups-high risk and low risk-in human genetic disease and complex genetic traits. However, it cannot handle uncertainties from genetic information. This chapter describes the fuzzy sigmoid membership-based MDR (FSMDR) method of epistasis detection. The algorithmic steps in FSMDR are also elaborated with simulated data generated from GAMETES and a real coronary artery disease patient epistasis data set obtained from the Wellcome Trust Case Control Consortium (WTCCC). Moreover, a belief degree-associated fuzzy MDR framework is also proposed for epistasis detection, which can overcome the uncertainties of MDR-based methods. This framework improves the detection efficiency. It works like fuzzy set-based MDR methods. Simulated epistasis data sets are used to compare different MDR-based methods. Belief degree-associated fuzzy MDR was shown to gives good results by taking into account the uncertainly of the high/low risk classification.
上位性是预测、分类和怀疑人类遗传疾病的一个挑战。已经开发出许多用于检测上位性的技术、方法和工具。多因子降维（MDR）是用于检测上位性的常用方法。它在人类遗传疾病和复杂遗传特征中使用两个类别组-高风险和低风险。然而，它不能处理遗传信息中的不确定性。本章描述了基于模糊 sigmoid 隶属度的上位性检测方法（FSMDR）。还详细阐述了 FSMDR 的算法步骤，这些步骤是使用 GAMETES 生成的模拟数据和从 Wellcome Trust Case Control Consortium（WTCCC）获得的真实冠状动脉疾病患者上位性数据集生成的。此外，还提出了一种用于上位性检测的置信度相关模糊 MDR 框架，该框架可以克服基于 MDR 方法的不确定性。该框架通过考虑高低风险分类的不确定性，提高了检测效率。它的工作原理类似于基于模糊集的 MDR 方法。使用模拟上位性数据集来比较不同的基于 MDR 的方法。置信度相关模糊 MDR 考虑到高低风险分类的不确定性，结果良好。
