What are antiviral compounds?

Antiviral compounds are a category of antimicrobial drugs used specially for treating viral infections by inhibiting the development of the viral pathogen inside the host cell.

How can we predict antivrial compounds using AVCpred?

On the submit page user can paste/upload the SDF files of the query molecules or also draw the chemical structure (using Draw tool) of the query molecule by using JME editor. On submission, AVCpred returns with predicted values on different viruses as initially selected by the user.

How were the prediction models built?

To develop QSAR models, we have calculated the different chemical descriptors (1D, 2D and 3D) and fingerprints by using PaDEL software (http://padel.nus.edu.sg/software/padeldescriptor/). We have developed separate QSAR models for each of the viruses and also a General model (26 viruses) by using SMOreg algorithm available in Weka machine learning package. Selected molecular descriptors calculated by the PaDEL were used as input. Feature selection was done using 'RemoveUseless' filter followed by ClassifierSubsetEval (attribute evaluator) with BestFirst (search method) module available in Weka package. In order to evaluate performance of our models, we used Pearson’s correlation coefficient (PCC). All models were evaluated using ten-fold cross validation technique.

What is the source of datasets used in this study?

We used the data from the ChEMBL (https://www.ebi.ac.uk/chembl/target/browser/). In this project,we selected we selected HIV, HCV, HHV, HBV (Virus specific datasets) and 26 other viruses (General dataset) targeted by 389, 467, 112, 124 and 1391 compounds respectively.

How is AVCpred useful to researchers?

Researchers can virtually screen and discover new antivirals (natural/synthetic) to extend the range of these compounds as well as design analogs of existing antivirals and predict their activity.