Browser compatibility:The functionality of VirulentPred 2.0 web-server is tested on the following web browsers:

OS Version Chrome Firefox Microsoft Edge Safari
Linux Ubuntu 18.04 81.0.4044.138 78.0.2 not tested n/a
MacOS Monterey 12.5 104.0.5112.101 103.0.2 not tested 15.6
Windows 7 104.0.5112.102 104.0 104.0.1293.70 not tested

VirulentPred 2.0 provides an easy to use platform to classify the bacterial proteins into virulent or non-virulent types. The screenshots displaying functioning of web-server (Figure 1) and standalone version (Figure 2) are shown below:

Figure 1. Screenshots displaying the functioning of VirulentPred 2.0 web-server. The prediction through web-server is broadly a four step process (A - D). Firstly, protein sequences (in FASTA format only) are submitted by a user (step A). The submitted sequences go to Que (step B). If enough resources are available, the sequences go under the processing for PSSM calculation and input file preparation for prediction by ML model (step C). Finally, after prediction through AutoGluon-based ML model, the output prediction results are available for download in the form of CSV file or Excel file (step D). The prediction results are also displayed as HTML table on the web page.



Figure 2. Screenshots displaying the functioning of VirulentPred 2.0 standalone. The prediction through standalone is a simple two step process (A - B). Firstly, protein sequences (in FASTA format only) are copied in the "virulentpred_2_0" directory. Single or multiple sequences are allowed to use with the standalone as depicted in the Step A. To run the prediction, a command i.e., "perl predict.pl sample_seq.fas" is applied within the "virulentpred_2_0" directory (through Linux terminal) as depicted in step B. The two steps involved in the prediction are: (i) calculation of PSSM profiles for protein sequences. (ii) proteins class (Virulent or Non-virulent) prediction through AutoGluon-based ML model. Finally, the predcition results are saved in the "CSV" and "Excel" file.