Table 1 Data and datasets.

From: RNA independent fragment partition method based on deep learning for RNA secondary structure prediction

Step

Dataset

Data group

Number of subsequences used/Number of RNAs in raw data

Number of subsequences (percentage)a

Pre-training

Training set (T1)

Group 3

17,927/105,370

14,342 (80%)

Validation set (V1)

  

3,585 (20%)

Trans-training

Training set (T2)

Group 2

3,635/24,863

2545 (70%)

Validation set (V2)

  

545 (15%)

Test

Test set (TS)

  

545 (15%)

Test set (TS’)

Group 1

128/1,003

128 (100%)

  1. aPercentage to the corresponding group (group 1,2 or 3).
Back to article page