Skip to main content

Table 1 Performance evaluation of tissue-conserved m6A sites of using different sequence strategy and algorithms

From: m6A-TCPred: a web server to predict tissue-conserved human m6A sites using machine learning approach

Sequence Strategy

Algorithm

Independent Testing

Sn

Sp

ACC

MCC

AUROC

NCP + ND

SVM

0.598

0.589

0.593

0.187

0.628

NB

0.655

0.465

0.560

0.123

0.585

GLM

0.587

0.581

0.584

0.169

0.621

NCP + EIIP

SVM

0.624

0.614

0.619

0.239

0.669

NB

0.707

0.469

0.585

0.186

0.636

GLM

0.604

0.619

0.612

0.224

0.660

EIIP + PseKNC

SVM

0.641

0.600

0.620

0.240

0.663

NB

0.921

0.188

0.555

0.162

0.635

GLM

0.604

0.606

0.605

0.210

0.648

  1. The SVM (support vector machine) represent binary classification method. NB refers to the naïve bayes classification method. GLM (generalized linear model) is linear regression model. The following sequence encoding strategy, NCP refers to the nucleotide chemical property [43]. ND is nucleotide density [55]. EIIP refers to electron–ion interaction pseudopotential (EIIP) [45]. PseKNC refers to Pseudo K-tuple nucleotide composition [56]