Вернуться к статье

Machine learning using multiple logistic regression for antimicrobial and hemolytic peptides prediction and their identification in large proteins

Table 1 - Sizes of sets used in ML for hemolytic activity prediction

Set

Split

Number of peptides

Hemolytic (390)

Training set: 80%

312

Test set: 20%

78

Non-hemolytic (390)

Training set: 80%

312

Test set: 20%

78

 

Training set: 80% (624)

Test set: 20% (156)

Sum: 780