PepFuncML - Train

Train Your Model

PepFuncML: A one-stop and comprehensive platform for small sample prediction of peptide efficacy and machine learning model construction.

Upload Your Dataset: The program only accepts data in CSV format. Please edit the file according to the EXAMPLE format.

Example

Select Sample Data Distribution:

only positive data positive and negative data

Data Augmentation: The program will calculate the distribution based on the length of protein fragments. For proteins with fewer occurrences and more than 10 amino acids, one amino acid will be mutated to a similar or similarly characteristic amino acid.

Yes No

Add Negative Data: Select the proportion of Positive and Negative Data

10:1

8:1

5:1

2:1

1:1

1:2

1:5

1:8

1:10

Partitioning the Test Set:

Select evaluation metric for model ranking:

Choose the folds number of the StratifiedKFold:

5 folds 10 folds

Optimize Model Parameters:
This article adopts Bayesian optimization for model parameter tuning to reduce the time required for parameter optimization. However, parameter tuning is time-consuming, please confirm whether you want to proceed with parameter optimization.

Yes No

Email address:

Train Your Model

Privacy Policy & Terms of Use