Bt Research 2025, Vol.16, No.2, 79-85 http://microbescipublisher.com/index.php/bt 81 reveal the action pathways and molecular mechanisms of toxins, and also improve the reliability of functional studies (Canzler et al., 2020; Subramanian et al., 2020; Shi et al., 2024). 4.2 Machine learning and bioinformatics for candidate prioritization Machine learning and bioinformatics tools have become important methods for the selection of candidate genes for Bt toxins. By integrating data at different levels such as genome, transcriptome and proteome, and combining techniques such as network analysis, graph neural network and similarity analysis, toxin genes with potential insecticidal activity can be efficiently predicted and screened (Subramanian et al., 2020). These methods enhance the efficiency of candidate screening and enable the mining of complex molecular interaction relationships, achieving data-driven functional prediction and target optimization (Athieniti and Spyrou, 2022; Jiang et al., 2025). 4.3 Structural modeling for predicting toxin-receptor interactions Structural modeling technology provides a powerful tool for predicting the binding mechanism between Bt toxins and insect receptors. Through methods such as homology modeling, molecular alignment, and molecular dynamics simulation, the binding sites and affiniances of toxin proteins and receptors can be predicted at the atomic level, which is helpful for screening and optimizing toxin molecules with new mechanisms of action (Figure 1) (Athieniti and Spyrou, 2022; Jiang et al., 2025). The combination of structural modeling, omics data and machine learning methods can achieve high-throughput screening from sequence to function. Figure 1 A workflow diagram from omics data generation to network-based analysis (Adopted from Jiang et al., 2025) Image caption: (A) Integration of multi-omics across the major omics layers (genomics, epigenomics, transcriptomics, proteomics, metabolomics, and phenomics); (B) The aggregation and archiving of multi-omics data into specialized databases, showcasing how these data repositories support standardization, preservation, and accessibility of large-scale biological data; (C) Multi-omics data extracted from databases are used to construct various biological networks (GRNs, PPI networks, MRNs, STNs, epigenetic networks, DTIs). The complex interactions in these networks reflect different biological activities; (D) The application of network-based analytical techniques to decipher the constructed networks. It includes methods like network feature selection methods and feature extraction methods (Adopted from Jiang et al., 2025) 5 Case Study: Application of HTS in Novel Bt Toxin Discovery 5.1 Study design: sample sources, library construction, and screening methods Samples in high-throughput screening studies of novel Bt toxins usually come from multiple Bt strains, environmental isolates, or mutants obtained through molecular evolution. Library construction includes toxin gene libraries based on genomes and metagenomes, as well as mutant libraries obtained through directed evolution
RkJQdWJsaXNoZXIy MjQ4ODYzNA==