banner

The following 2 datasets were derived from the APD database. AMPs in the database were shown to inhibit against gram+ and gram- bacteria

Positive_training_set (1500)

Positive_testing_set (139)



The following 6 datasets are negative datasets that were used to train and test the model. The datasets are short sequences that do not contain AMP characteristics.

Negative training set (1500), NOAMP1

Negative testing set (139), NOAMP1

Negative training set (1500), NOAMP2

Negative testing set (139), NOAMP2

Negative training set (1500), NOAMP3

Negative testing set (139), NOAMP3



The following 2 datasets are peptides from the Citrus genome that were predicted positive and negative by AGRAMP and these datasets were used
for the comparison with other AMP prediction models.

Testing dataset (200), AGRAMP AMP

Testing dataset (200), AGRAMP NOAMP



The dataset of 500 Citrus genome peptides that were predicted positive by AGRAMP

AGRAMP_AMP