NetPicoRNA

NetPicoRNA predicts cleavage sites of picornaviral proteases 2Apro and 3Cpro in viral and cellular proteins using neural networks and a surface exposure prediction algorithm to identify sequence and structural determinants of protease specificity.

Key Features:

• Neural Network Algorithms: Uses neural network models trained on known cleavage site sequences to detect sequence patterns associated with 2Apro and 3Cpro specificity.
• Graphical Visualization Techniques: Provides graphical visualization of predictive outputs and sequence features to represent cleavage site predictions.
• Surface Exposure Prediction Algorithm: Integrates a surface exposure prediction algorithm to differentiate exposed from buried residues and refine cleavage site classification.
• Experimental Validation of Cellular Sites: Predicts experimentally determined cleavage sites in cellular proteins in addition to viral polyprotein maturation sites.
• Database Error Detection: Identifies and reports errors in biological databases related to reported cleavage-site annotations.

Scientific Applications:

• Viral polyprotein processing: Analysis of maturation cleavages mediated by picornaviral proteinases 2Apro and 3Cpro within viral polyproteins.
• Host target identification: Prediction of potential cellular protein targets of picornaviral proteases for experimental follow-up.
• Viral–host interaction studies: Support for investigations into how picornaviruses manipulate host cellular machinery via proteolytic cleavage.
• Annotation quality control: Contribution to improving the accuracy of cleavage-site annotations in biological databases.

Methodology:

Neural networks are trained on known picornaviral protease cleavage site data and predictions are combined with a surface exposure prediction algorithm to consider both sequence and structural features.