


About us Profile Management Scientific Advisory Board Strategic alliances Partners Top articles Publications Platform technologies Nested Chemical Library The concept of Nested Chemical Library Chemical Validation Library Master library Hit finding Allosteric library Allosteric library Developing novel allosteric inhibitors Masterkey The Masterkey concept Masterkeys from the Nested Chemical Library Pharmacophore modeling Scaffold hopping The concept of scaffold hopping Scaffold hopping based lead finding Lead optimization - ADMET Lead generation Lead optimization Experimental ADMET parameters KinaTor Services Hit and lead finding Kinase inhibitor library In vitro biology Lead and ADMET optimization Target fishing Predictive QSAR Custom synthesis Process development Drug discovery Drug discovery programs Epilepsy EGFR HIV Influenza Neurogenic inflammation Tuberculosis Tumor stem cells Contact us

Predictive QSAR

We have developed a predictive QSPAR system using our in house developed DHELPHO software. We have developed more than a hundred QSPAR and pharmacophore models for a big series of kinases and ADMET parameters.

Our QSAR Machine service is accessible for registered users.

The toll-free models can estimate physico-chemical and toxicological parameters. Commercial users have access to our other (ADME and enzyme-inhibition) models and also can initiate model-building from their own data.

Using our available or user-defined models we give you the opportunity for estimating the biological effects of a given structure or for searching in the whole Vichem chemical database.

You can calculate selectivity profile in our large database (approx. 11 000 000 structures) or estimate pharmacological parameters besides the biological effects using multiple concurrent models.

Vichem offers to perform chemical synthesis of the selected compounds - on a special contractual basis.

For targets we do not have available models yet Vichem develops the best externally validated possible QSAR model obtainable from Your data. You provide your data set (of any kind) - we deliver the model.

Our Predictive QSAR modeling algorithms (PLS, ANN, ADABOOST etc.) are optimized for highest predictive ability on external compound series. The reliability of the predictions are measured by predictions on external validation sets and experimentally.