首页|A hybrid feature selection method for predicting lysine malonylation sites in proteins via machine learning
A hybrid feature selection method for predicting lysine malonylation sites in proteins via machine learning
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NSTL
Elsevier
Providing efficient methods for analyzing medical data is one of the important needs of modern biological sci-ences. For this, in this paper, a new feature selection method is introduced using the combination of several feature selection methods. At first, the algorithms of EAAC, EGAAC, PKa, TF-IDF, TF-CRF, and PSSM are expressed, which are among the well-known methods for feature extraction, and then three proposed models are provided that are the combinations of these algorithms. The proposed method has been implemented on three lysine malonylation datasets of M. musculus, H. sapiens, and E. coli, and also several machine learning methods have been used to categorization the data. Finally, to show the efficiency of the proposed method, some important parameters have been calculated and compared with other feature extraction methods. Furthermore, the results have been compared with several well-known articles and the results have been reported tabularly and graphically.
Malonylation feature extraction feature p selection machine learning deep learningSUCCINYLATION
Rajabiun, Hananeh、MohammadHoseini, Mahdis、Zarezadeh, Hadi、Delkhosh, Mehdi
NSTL
Elsevier
