REPOSITORY QDB GUIDELINES CITING BLOG

Oja, M.; Sild, S.; Maran, U. Logistic Classification Models for pH-Permeability Profile: Predicting Permeability Classes for the Biopharmaceutical Classification System. J. Chem. Inf. Model. 2019, 59, 2442-2455.

QsarDB Repository

Oja, M.; Sild, S.; Maran, U. Logistic Classification Models for pH-Permeability Profile: Predicting Permeability Classes for the Biopharmaceutical Classification System. J. Chem. Inf. Model. 2019, 59, 2442-2455.

QDB archive DOI: 10.15152/QDB.206   DOWNLOAD

QsarDB content

Property Class_pH3: Class of membrane permeability at pH 3 i

Compounds: 238 | Models: 2 | Predictions: 8

M1: Model at pH 3 with hydrophobicity descriptor

Logistic regression (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
Training set training 67 0.910
Validation set external validation 111 0.820
External validation set external validation 60 0.900
FDA reference drugs for BCS permeability class i testing 32 N/A
M7: Model at pH 3 with theoretical molecular descriptors

Logistic regression (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
Training set training 67 0.881
Validation set external validation 111 0.793
External validation set external validation 60 0.833
FDA reference drugs for BCS permeability class i testing 32 N/A

Property Class_pH5: Class of membrane permeability at pH 5 i

Compounds: 238 | Models: 2 | Predictions: 8

M2: Model at pH 5 with hydrophobicity descriptor

Logistic regression (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
Training set training 89 0.843
Validation set external validation 89 0.730
External validation set external validation 60 0.800
FDA reference drugs for BCS permeability class i testing 32 N/A
M8: Model at pH 5 with theoretical molecular descriptors

Logistic regression (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
Training set training 89 0.764
Validation set external validation 89 0.753
External validation set external validation 60 0.733
FDA reference drugs for BCS permeability class i testing 32 N/A

Property Class_pH7.4: Class of membrane permeability at pH 7.4 i

Compounds: 238 | Models: 2 | Predictions: 8

M3: Model at pH 7.4 with hydrophobicity descriptor

Logistic regression (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
Training set training 134 0.858
Validation set external validation 44 0.795
External validation set external validation 60 0.883
FDA reference drugs for BCS permeability class i testing 32 N/A
M9: Model at pH 7.4 with theoretical molecular descriptors

Logistic regression (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
Training set training 134 0.828
Validation set external validation 44 0.886
External validation set external validation 60 0.717
FDA reference drugs for BCS permeability class i testing 32 N/A

Property Class_pH9: Class of membrane permeability at pH 9 i

Compounds: 238 | Models: 2 | Predictions: 8

M4: Model at pH 9 with hydrophobicity descriptor

Logistic regression (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
Training set training 134 0.858
Validation set external validation 44 0.864
External validation set external validation 60 0.717
FDA reference drugs for BCS permeability class i testing 32 N/A
M10: Model at pH 9 with theoretical molecular descriptors

Logistic regression (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
Training set training 134 0.813
Validation set external validation 44 0.909
External validation set external validation 60 0.767
FDA reference drugs for BCS permeability class i testing 32 N/A

Property Class_highest: Class of highest membrane permeability for pH range 3 to 9 i

Compounds: 238 | Models: 2 | Predictions: 8

M5: Model for highest membrane permeability with hydrophobicity descriptor

Logistic regression (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
Training set training 107 0.850
Validation set external validation 71 0.887
External validation set external validation 60 0.733
FDA reference drugs for BCS permeability class i testing 32 N/A
M11: Model for highest membrane permeability with theoretical molecular descriptors

Logistic regression (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
Training set training 107 0.850
Validation set external validation 71 0.817
External validation set external validation 60 0.800
FDA reference drugs for BCS permeability class i testing 32 N/A

Property Class_logPo: Class of intrinsic membrane permeability i

Compounds: 238 | Models: 2 | Predictions: 8

M6: Model for intrinsic membrane permeability (logPo) with hydrophobicity descriptor

Logistic regression (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
Training set training 91 0.857
Validation set external validation 87 0.897
External validation set external validation 60 0.767
FDA reference drugs for BCS permeability class i testing 32 N/A
M12: Model for intrinsic membrane permeability (logPo) with theoretical molecular descriptors

Logistic regression (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
Training set training 91 0.813
Validation set external validation 87 0.851
External validation set external validation 60 0.783
FDA reference drugs for BCS permeability class i testing 32 N/A

Property Class_FDA: BCS permeability class for FDA reference drugs i

Compounds: 32 | Models: 2 | Predictions: 2

Fig5_1: Decision tree based on models with hydrophobicity descriptors

Ensemble model (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
FDA reference drugs for BCS permeability class training 32 0.813
Fig5_2: Decision tree based on the models with theoretical molecular descriptors

Ensemble model (classification)

Open in:QDB Explorer QDB Predictor

Name Type n Accuracy
FDA reference drugs for BCS permeability class training 32 0.906

Citing

When using this QDB archive, please cite (see details) it together with the original article:

Metadata

Show full item record

Title: Oja, M.; Sild, S.; Maran, U. Logistic Classification Models for pH-Permeability Profile: Predicting Permeability Classes for the Biopharmaceutical Classification System. J. Chem. Inf. Model. 2019, 59, 2442-2455.
Abstract: Permeability is used to describe and evaluate the absorption of drug substances in the human gastrointestinal tract (GIT). Permeability is largely dependent on fluctuating pH that causes the ionization of drug substances and also influences regional absorption in the GIT. Therefore, classification models that characterize permeability at wide range of pH-s were derived in the current study. For this, drug substances were described with six data series that were measured with a parallel artificial membrane permeability assay (PAMPA), including a permeability profile at four pH-s (3, 5, 7.4 and 9), and the highest and intrinsic membrane permeability. Logistic regression classification models were developed and compared by using two distinct sets of descriptors: 1) hydrophobicity descriptor, the logarithm of the octanol-water partition (logPow) or distribution (logD) coefficient, and 2) theoretical molecular descriptors. In both cases, models have good classification and descriptive capabilities for training set (accuracy: 0.76 to 0.91). Triple-validation with three sets of drug substances shows good prediction capability for all models: validation set (accuracy: 0.73 to 0.91), external validation set (accuracy: 0.72 to 0.9) and the permeability classes of FDA reference drugs for the biopharmaceutical classification system (BCS) (accuracy: 0.72 to 0.88). The identification of BCS permeability classes was further improved with decision trees that consolidated predictions from models with each descriptor type. These decision trees have higher confidence and accuracy (0.91 for theoretical molecular descriptors and 0.81 for hydrophobicity descriptors) than the individual models in assigning drug substances into BCS permeability classes. A detailed analysis of classification models and related decision trees suggests that they are suitable for predicting classes of permeability for passively transported drug substances, including specifically within the BCS framework.
URI: http://hdl.handle.net/10967/206
http://dx.doi.org/10.15152/QDB.206
Date: 2019-02-21


Files in this item

Name Description Format Size View
2019JCIM.zip Classification models and decision tree for membrane permeability application/x-zip 125.9Kb View/Open
Files associated with this item are distributed
under Creative Commons license.

This item appears in the following Collection(s)

Show full item record