NetCorona

NetCorona predicts 3C-like protease (3CLpro) cleavage sites in amino acid sequences to identify processing sites in coronavirus polyproteins and potential host substrates.

Key Features:

• Data-Driven Analysis: Uses sequence alignment data from seven fully sequenced coronaviruses to identify conserved 3CLpro cleavage sites.
• Neural Network Model: Employs a neural network trained to recognize sequence patterns associated with 3CLpro cleavage across coronavirus genomes.
• Sequence-Context and Evolutionary Features: Incorporates features derived from local sequence context and evolutionary conservation to inform predictions.
• Performance Metrics: Reports a sensitivity of 87.0% and a specificity of 99.0% on benchmark datasets.
• Viral and Host Substrate Scope: Predicts cleavage sites in viral polyproteins and candidate host proteins including CFTR, CREB-RP, OCT-1, and components of the ubiquitin pathway.

Scientific Applications:

• Pathogenesis Mapping: Maps 3CLpro processing sites to support studies of coronavirus replication and pathogenesis.
• Therapeutic Target Identification: Identifies candidate cleavage sites to prioritize targets for 3CLpro inhibitor development and other therapeutic strategies.
• Host Protein Impact Assessment: Identifies potential cleavage of host proteins to investigate how coronaviruses manipulate host cellular machinery and to prioritize sites for experimental validation.

Methodology:

Integrates sequence alignment data from seven coronaviruses with a neural network trained to predict 3CLpro cleavage sites using features derived from sequence context and evolutionary conservation, with predictions validated using benchmark datasets.