Scholarlink

<p>The identification of protein-coding&#xa0;regions using the three-base periodicity property&#xa0;is a challenging task&#xa0;in the field of bioinformatics. Many digital signal processing (DSP)-based techniques have been used to extract the period-3 component. An effective DSP-based technique must satisfy three crucial requirements: being independent of window length, having an adaptive frequency spectrum, and having a center frequency of <i>f</i>/3 to identify the protein-coding regions (exons). The sinusoidal-assisted variational mode decomposition (SAVMD) resolves the issue to some extent&#xa0;by utilizing a sinusoidal-assisted framework with fixed parameters of VMD. However, the performance of the SAVMD has been reduced because of the fixed parameters used in VMD. Further, the fixed parameters have not been suitable for different DNA sequences to properly identify the exons. Therefore, a novel optimized SAVMD (OSAVMD) method is proposed based on different swarm optimization algorithms for the accurate identification of protein-coding regions. Here the least energy loss coefficient is used as the fitness function to optimize the parameters (<i>α</i> and <i>K</i>) of the standard variational mode decomposition to achieve better accuracy in the prediction of the exons. In order to evaluate the efficacy of the proposed method compared to existing methods, some critical performance parameters are used at the nucleotide level for benchmark gene sets like C. elegans, Homosapiens, Mus musculus, etc. The results reveal that the OSAVMD technique achieves superior performance than the other existing methods.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Identification of protein-coding regions using optimized sinusoidal assisted variational mode decomposition based on swarm optimization algorithm

  • 期刊论文

  • 发表日期：

    2025年4月1日

查看全文
  • K. Jayasree,
  • Malaya Kumar Hota

摘要

The identification of protein-coding regions using the three-base periodicity property is a challenging task in the field of bioinformatics. Many digital signal processing (DSP)-based techniques have been used to extract the period-3 component. An effective DSP-based technique must satisfy three crucial requirements: being independent of window length, having an adaptive frequency spectrum, and having a center frequency of f/3 to identify the protein-coding regions (exons). The sinusoidal-assisted variational mode decomposition (SAVMD) resolves the issue to some extent by utilizing a sinusoidal-assisted framework with fixed parameters of VMD. However, the performance of the SAVMD has been reduced because of the fixed parameters used in VMD. Further, the fixed parameters have not been suitable for different DNA sequences to properly identify the exons. Therefore, a novel optimized SAVMD (OSAVMD) method is proposed based on different swarm optimization algorithms for the accurate identification of protein-coding regions. Here the least energy loss coefficient is used as the fitness function to optimize the parameters (α and K) of the standard variational mode decomposition to achieve better accuracy in the prediction of the exons. In order to evaluate the efficacy of the proposed method compared to existing methods, some critical performance parameters are used at the nucleotide level for benchmark gene sets like C. elegans, Homosapiens, Mus musculus, etc. The results reveal that the OSAVMD technique achieves superior performance than the other existing methods.