Bt Research 2024, Vol.15, No.5, 240-247 http://microbescipublisher.com/index.php/bt 244 2020). Another study utilized phage-assisted continuous evolution to develop Cry1Ac variants that bind to a cadherin-like receptor in Trichoplusia ni, overcoming resistance and enhancing the lethality of the toxin (Badran et al., 2016). Figure 2 The detailed pipeline implemented in CryProcessor (Adopted from Shikov et al., 2020) 5.3 Predictive models for toxin-target specificity Predictive models have been developed to understand the specificity of Bt toxins to their targets. For example, CRISPR-mediated gene knockouts were used to evaluate the role of five candidate Bt toxin receptors in Spodoptera exigua. The study identified SeABCC2 as a major receptor mediating the toxicity of Cry1Ac and Cry1Fa, providing insights into the genetic mechanisms of resistance and aiding in the development of more effective monitoring and management strategies (Huang et al., 2020; 2023). The molecular characterization of Cry toxin receptor-like genes in Galleria mellonella revealed that the suppression of these receptors significantly reduced larval sensitivity to Cry1AcF toxin, offering valuable information for future research on Bt resistance (Dutta et al., 2022). 6 Challenges in Applying AI to Bt Toxin-Insect Studies 6.1 Data limitations and model accuracy One of the primary challenges in applying AI to Bt toxin-insect studies is the limitation and quality of available data. Accurate AI models require extensive and high-quality datasets to train effectively. However, the current datasets on Bt toxins and insect interactions are often fragmented and incomplete. For instance, the CryProcessor tool was developed to address the limitations of existing tools like BtToxin_scanner, which suffer from limited query size and outdated databases (Shikov et al., 2020). The variability in data quality can significantly impact the
RkJQdWJsaXNoZXIy MjQ4ODYzNA==