SAWRPI: A Stacking Ensemble Framework With Adaptive Weight for Predicting ncRNA-Protein Interactions Using Sequence Information

• DOI: 10.3389/fgene.2022.839540
• Publication Date: 2022-02-28T06:30:45Z
• Authors: Zhong-Hao Ren; Chang-Qing Yu; Li-Ping Li; Zhu-Hong You; Yong-Jian Guan; Yue-Chao Li; Jie Pan
• Keywords: Robustness; Feature (linguistics)

Non-coding RNAs (ncRNAs) take essential effects on biological processes, like gene regulation. One critical way of ncRNA executing functions is interactions between and RNA binding proteins (RBPs). Identifying proteins, involving ncRNA-protein interactions, can well understand the function ncRNA. Many high-throughput experiment have been applied to recognize interactions. As a consequence these approaches are time- labor-consuming, currently, great number computational methods developed improve advance research. However, may be not available all particularly processing new proteins. Additionally, most them cannot process with long sequence. In this work, method SAWRPI proposed make prediction through sequence information. More specifically, raw features protein firstly extracted k-mer sparse matrix SVD reduction learning nucleic acid symbols by natural language local fusion strategy, respectively. Then, classify easily, Hilbert Transformation exploited transform feature data space. Finally, stacking ensemble strategy adopted learn high-level abstraction automatically generate final results. To confirm robustness stability, three different datasets containing two kinds utilized. comparison state-of-the-art other results classifying or extracting strategies, achieved high performance datasets, lncRNA-protein Upon our finding, trustworthy, robust, yet simple used as beneficial supplement task predicting