Genomic splice site prediction algorithm based on nucleotide sequence pattern for RNA viruses.

Area of Science:

• Bioinformatics
• Virology
• Computational Biology

Background:

• RNA virus splice site prediction faces challenges due to limited strains and high mutation rates.
• Conventional predictors struggle with diversified sequence patterns and insufficient training data.

Purpose of the Study:

• To develop a novel algorithm, Genomic Splice Site Prediction (GSSP), for accurate splice site prediction in RNA viruses.
• To address the limitations of conventional methods in handling RNA virus sequence variability.

Main Methods:

• The GSSP algorithm characterizes nucleotide interdependencies using eigen-patterns identified via sequence pattern mining.
• A cross-species strategy was implemented to overcome insufficient training data from limited viral strains.
• The algorithm was tested on five RNA species within the Orthomyxoviruses family.

Main Results:

• GSSP demonstrated superior performance compared to two conventional methods.
• High prediction accuracy was achieved: >99% sensitivity and >94% specificity for donor sites.
• >96% sensitivity and >92% specificity for acceptor sites were recorded.

Conclusions:

• The GSSP algorithm is effective and robust for splice site prediction in RNA viruses.
• The eigen-pattern characterization and cross-species strategy successfully mitigate challenges posed by viral mutation and data scarcity.