COVID-19 vs influenza viruses: A cockroach optimized deep neural network classification approach

Area of Science:

• Virology
• Computational Biology
• Artificial Intelligence

Background:

• Viral pathogenesis, including COVID-19, is heavily influenced by host cell receptor interactions.
• Understanding these interactions is crucial for diagnosing and differentiating viral infections.
• Accurate and rapid diagnostic tools are essential for managing outbreaks.

Purpose of the Study:

• To develop a deep neural network for detecting COVID-19.
• To differentiate COVID-19 from influenza types A, B, and C using genomic sequence data.
• To optimize the deep neural network architecture using a cockroach optimization algorithm.

Main Methods:

• Utilized genomic sequences of COVID-19 and influenza A, B, and C.
• Employed a deep neural network architecture inspired by a cockroach optimization algorithm.
• Trained and tested the model using 594 unique genome sequences.

Main Results:

• Achieved 99% overall accuracy in classifying viral sequences.
• Successfully differentiated between COVID-19 and influenza types A, B, and C.
• Demonstrated the efficacy of the cockroach optimization algorithm in enhancing deep network performance.

Conclusions:

• The proposed cockroach-optimized deep neural network is a highly accurate tool for COVID-19 detection and differentiation from influenza.
• This approach offers a promising computational method for viral diagnostics.
• Further research can explore this optimization technique for other viral pathogens.