Analyzes overview
| Tool | Description |
|---|---|
| Binding Site | Correctly predicting the binding site on earlier stages of the drug discovery process may have a crucial impact on the primary of your project. By our tool - KALASANTY as well as DeepPocket you may investigate the surface of your protein target and predict druggable pockets that can be later used to design active molecules. For this purpose, we implemented a fully convolutional 3D neural network capable of binding site segmentation which has high prediction accuracy and provides intuitive representations of the results, which makes it easy to incorporate into your drug discovery projects. Second implemented a DeepPocket (Aggarwal et al. 2021) utilizes 3D convolutional neural networks for the rescoring of pockets identified by a geometry-based software called Fpocket (Le Guilloux et al. 2009). Both methods complement each other perfectly and predictions of your binding site(s) location(s) should be available in minutes, which could be further used inside the SilDrug Service or visualized by other common software like PyMOL or VMD. |
| Docking | Our web available in silico high throughput molecular screening system is based on the open-source VINA docking engine, which has been found as a strong competitor against other, high-end docking programs distributed commercially. |
| Rescoring | Classical methods supporting the drug discovery process have reached a plateau in their performance in virtual screening and binding affinity prediction, thus we developed and investigated a set of machine-learning methods that have shown great promise in the field: PLEC is a method to represent ligand-receptor complexes based on local atomic environments, which allows a very fast and accurate description of the interaction of atoms in complexes, which can become the basis for training and optimization of prediction models, e.g. binding affinity values. Pafnucy is a first-in-class convolutional neural network to predict the affinity of compounds based on their three-dimensional complexes with desired targets. This network is able not only to distinguish active compounds from inactive, but more importantly provides affinity value allowing for ranking list development. RF-Score-VS is our new machine-learning scoring function for structure-based binding affinity prediction of protein-ligand complexes, which has been trained on 15 426 active and 893 897 inactive molecules docked to a set of 102 targets and is ready to support your drug development project and evaluate your docking/screening results. |
| Pharmacophore | DeCAF is a solution for comparing and assessing compounds based on pharmacophore features. DeCAF offers a unique measure of compounds similarity based on their physicochemical characteristics, abstracting from their chemical structure. DeCAF allows large-scale comparisons of low-molecular weight compounds, creating extensive pharmacophore alignments, and thus building models for assessing the activity of chemical molecules without knowing the structure of their complexes with the receptor. |
| ECBD | Using SilDrug Service you may find useful our implementation of the ECBD (European Chemical Biology Database) which collects experimental results from biological screening programs performed within the EU-OPENSCREEN platform and is hosted by the Czech EU-OPENSCREEN partner site at the Institute of Molecular Genetics (IMG) in Prague belonging to CZ-OPENSCREEN. There are 3 chemical libraries: bioactive, diverse, and academic. The bioactive library contains 2464 compounds selected with a strong emphasis on target coverage and selectivity. The diverse library contains approx. 100 000 compounds with unbiased chemical diversity, designed by five academic computational chemistry groups. In the academic library, compounds are collected from European chemists and their number will be growing over the next few years. Structures available in the our implementation of ECBD may be downloaded and directly used in the SliDrug Service projects (e.g. docking, pharmacophore modeling). |