##PPISP-XGBoost
Prediction of protein-protein interaction sites through eXtreme gradient boosting with kernel principal component analysis
###PPISP-XGBoost uses the following dependencies:
- Python 3.6
- numpy
- scipy
- scikit-learn
- pandas
###Guiding principles:
**The dataset file contains five datasets, among which dset72_fasta, dset164_fasta, dset186_fasta, Dset_448_fasta, Dset_355_fasta.
**Feature extraction：
- PseAAC.py is the implementation of PseAAC.
- PsePSSM.m is the implementation of PsePSSM.
- pssm_example.example, pssm_slide.m and PSSM_creat.m are the implementation of PsePSSM.
- ASA_example.spd33 is the implementation of ASA.
- Hy_index_exampie.index is the implementation of hydrophilic index.
**Feature_selection:
- FA_selection is the implementation of FA.
- ICA_selection is the implementation of ICA.
- KPCA_selection is the implementation of KPCA.
- LASSO_selection is the implementation of LASSO.
- LR_selection is the implementation of LR.
- MI_selection is the implementation of MI.
- PCA_selection is the implementation of PCA.
- SE_selection is the implementation of SE.
**SMOTE:
- SMOTE.R is the implementation of SMOTE.
**Classifier:
- AdaBoost_classifier.py is the implementation of Adaboost.
- DT_classifier.py is the implementation of DT.
- GBDT_classifier.py is the implementation of GBDT.
- KNN_classifier.py is the implementation of KNN.
- RF_classifier.py is the implementation of RF.
- SVM_classifier.py is the implementation of SVM.
- XGBoost_classifier.py is the implementation of XGBoost.
- MLP_classifier.py is the implementation of MLP.
- NB_classifier.py is the implementation of NB.