A Novel Feature Fusion Method Based on RSCU and K-mer to Classify the SARS-Cov-2 DOI Creative Commons
Fuqiang Ye, Jianhua Zhou, Hao Zhang

и другие.

Academic Journal of Science and Technology, Год журнала: 2024, Номер 13(3), С. 334 - 342

Опубликована: Дек. 26, 2024

The SARS-Cov-2 virus exhibits a high mutation rate, which makes the prediction and classification of its genetic evolution variation trends highly significant. Accurate methods not only contribute to epidemiological studies virus, but also play crucial role in vaccine development antiviral drug discovery. This study aims systematically evaluate accuracy generalization capability RSCU (Relative Synonymous Codon Usage) K-mer encoding techniques SARS-CoV-2 genome. We extracted genomic data from two major variants, Alpha Beta, applied Support Vector Machine (SVM) algorithm train assess impact different feature on performance. Furthermore, we introduce novel multi-feature fusion method, KRSCU, combines sequence position information with synonymous codon compositions RSCU. method effectively captures subtle differences data, significantly improving both model. Experimental results demonstrate that KRSCU outperforms traditional single-feature approaches subtype tasks. Our research offers new insights into analysis, potential applications viral monitoring.

Язык: Английский

A Novel Feature Fusion Method Based on RSCU and K-mer to Classify the SARS-Cov-2 DOI Creative Commons
Fuqiang Ye, Jianhua Zhou, Hao Zhang

и другие.

Academic Journal of Science and Technology, Год журнала: 2024, Номер 13(3), С. 334 - 342

Опубликована: Дек. 26, 2024

The SARS-Cov-2 virus exhibits a high mutation rate, which makes the prediction and classification of its genetic evolution variation trends highly significant. Accurate methods not only contribute to epidemiological studies virus, but also play crucial role in vaccine development antiviral drug discovery. This study aims systematically evaluate accuracy generalization capability RSCU (Relative Synonymous Codon Usage) K-mer encoding techniques SARS-CoV-2 genome. We extracted genomic data from two major variants, Alpha Beta, applied Support Vector Machine (SVM) algorithm train assess impact different feature on performance. Furthermore, we introduce novel multi-feature fusion method, KRSCU, combines sequence position information with synonymous codon compositions RSCU. method effectively captures subtle differences data, significantly improving both model. Experimental results demonstrate that KRSCU outperforms traditional single-feature approaches subtype tasks. Our research offers new insights into analysis, potential applications viral monitoring.

Язык: Английский

Процитировано

0