Skip to main content
. 2021 Mar 22;12:630379. doi: 10.3389/fgene.2021.630379

Figure 1.

Figure 1

The MEL-MP workflow. The four sub-models correspond to four types of extracted features. Seq − BiLSTM denotes that the bidirectional long short-term memory neural network (BiLSTM) model is selected for primary sequence features. PSSM − RF means that the random forest (RF) classifier is used for PSSM evolution information. SS − ARF denotes that the classifier based on an auto-encoder and RF is applied for secondary structure features. AA − MLP means that the multilayer perceptron (MLP) model is applied for physical and chemical features. Finally, logistic regression (LR) integrates the outputs of all models.