SMMPPI is a machine learning based tool to predict small molecule based modulators of Protein-Protein Interactions. It involves two-stages: At First, it predicts the compounds with potential to act as PPI modulators. In the next stage, PPI family specific predictors can be used to predict modulators for 11 clinically important PPI familes.
DATASETS
-
General PPIMs Prediction (Stage I) Datasets
Data clustered with Tanimoto Data clustered with Tanimoto Data clustered with Tanimoto Score cutoff =0.9 Score cutoff =0.8 Score cutoff =0.7
Random-Split
3:1 Train-Test
Train Set Test Set Train Set Test Set Train Set Test Set
1:1 Train-Test
Train Set Test Set Train Set Test Set Train Set Test Set
AVE-Split
3:1 Train-Test
Train Set Test Set Train Set Test Set Train Set Test Set
1:1 Train-Test
Train Set Test Set Train Set Test Set Train Set Test Set
Realistic-Split
3:1 Train-Test
Train Set Test Set Train Set Test Set Train Set Test Set
1:1 Train-Test
Train Set Test Set Train Set Test Set Train Set Test Set
PPI Class Specific (Stage II) Datasets
1. Bromodomain_Histone Datasets
Train Set Test Set
2. BCL2-Like_BAX_BAK_comb Datasets
Train Set Test Set
3. LEDGF_IN Datasets
Train Set Test Set
4. LFA_ICAM Datasets
Train Set Test Set
5. MDM2-Like_P53 Datasets
Train Set Test Set
6. RAS_SOS1 Datasets
Train Set Test Set
7. XIAP_Smac Datasets
Train Set Test Set
8. RBD_ACE2 Datasets
Train Set
9. WDR5_MLL1 Datasets
Train Set
10. KEAP1_NRF2 Datasets
Train Set
11. CD4_gp120 Datasets
Train Set
Supplementary Data Sheet