Each file should contain the following set of data.
In the first column: the id numbers of all compounds included in the screened collection.
In each of the subsequent columns: the ranking of each compound
according to the corresponding virtual screening method.
You can add as many methods -and columns- as you want.
After uploading the file, wait for the application to display the results.
You can then set the file prefix in the
field and click
The output will be the list of all compounds sorted by the Consensus ranking method along
with their corresponding scores.
For the DYRK1a, GSK3b and CDK5 datasets please cite: "A facile consensus ranking approach enhances
virtual screening robustness and identifies a cell-active DYRK1a inhibitor"
For the CK1 dataset please cite: "Combined virtual and experimental screening for CK1 inhibitors identifies
a modulator of p53 and reveals important aspects of in silico screening performance"
International Journal of Molecular Sciences
, 18, 2102.
The method is based on a conceptual framework for exploiting the orthogonality aspect that is
present among the various available structure-based and ligand-based virtual screening tools.
The consensus scheme can utilize as many orthogonal primary VS methods as desirable.
Results of each individual method are averaged and normalized. Subsequently, a logarithmic transformation is applied
so that compounds that attain top ranks within individual screens are favored upon incorporation in consensus ranking.
The consensus ranking protocol enables considerable improvement of screening performance stability over individual VS approaches.