1,099
Views
2
CrossRef citations to date
0
Altmetric
Articles

A novel method for predicting RNA-interacting residues in proteins using a combination of feature-based and sequence template-based methods

, , , &
Pages 1138-1149 | Received 27 Nov 2018, Accepted 23 Apr 2019, Published online: 23 Jul 2019

Figures & data

Figure 1. Prediction performance for different sizes of sequence pattern on validation set RB86.

Figure 1. Prediction performance for different sizes of sequence pattern on validation set RB86.

Table 1. The one-hot binary encoding for 20 types of amino acids.

Figure 2. An example of adjustment procedure.

Figure 2. An example of adjustment procedure.

Figure 3. The flowchart of our proposed prediction method.

Figure 3. The flowchart of our proposed prediction method.

Figure 4. MCC and F-score with respect to different sizes of adjustment windows on validation set RB86.

Figure 4. MCC and F-score with respect to different sizes of adjustment windows on validation set RB86.

Table 2. Prediction performance of single and ensemble predictors on validation set RB86.

Figure 5. The proportion of predicted interacting residues by different prediction methods.

Figure 5. The proportion of predicted interacting residues by different prediction methods.

Table 3. Prediction performance comparison on testing set RB44.

Figure 6. A representative example result of RNA-interacting residue prediction on 3izv by feature-based predictor without the adjustment procedure (a), feature-based predictor with the adjustment procedure (b) and a combination of feature-based and sequence template-based predictors (c).

Figure 6. A representative example result of RNA-interacting residue prediction on 3izv by feature-based predictor without the adjustment procedure (a), feature-based predictor with the adjustment procedure (b) and a combination of feature-based and sequence template-based predictors (c).