Sequence-based prediction of protein–protein interaction using auto-feature engineering of RNN-based model

• Published in: Research on biomedical engineering 2023-03, Vol.39 (1), p.259-272
• Main Authors: Mewara, Bhawna, Lalwani, Soniya
• Format: Article
• Language: English
• Subjects: Biomaterials; Biomedical Engineering and Bioengineering; Biomedical Engineering/Biotechnology; Engineering; Original Article
• ISSN: 2446-4740; 2446-4740
• DOI: 10.1007/s42600-023-00273-z

Purpose The intricate language of eukaryotic gene articulation remains deficiently comprehended. Notwithstanding the significance recommended by numerous protein variations genuinely related to human infection, virtually all such variations have obscure systems, for instance, protein–protein connections (PPIs). Method This paper addresses the aforementioned challenge by employing a progressive recurrent neural network-based architecture as a prediction model with the aim of enhancing performance. Result This RNN-based architecture is carefully designed with suitable parameter settings so as to cover the maximum possible latent features in a more refined way and applied some techniques to avoid over-fitting issues. The efficiency of the proposed work is evaluated using two intra-species and two inter-species datasets. Also, the proposed approach has outperformed when compared with other state-of-the-art methods using Bonferroni post hoc analysis for all the considered datasets. Conclusion The promising outcomes and the comparison with other advanced methods of the proposed model advocate that auto-feature engineering and layer-wise abstraction helps to learn the essential features of protein pairs.