A dataset of gene expression from Spodoptera frugiperda, a highly generalist pest moth, was used to understand how gene regulation is related to larval host plant preference. Transcriptomic data of corn and rice strains of S. frugiperda larvae, reared on different diets, were analysed with three different approaches of gene network inference, namely co-expression, weighted co-expression and Bayesian networks, since each methodology provides a different visualization of the data. Using these approaches, it was possible to identify two loosely interconnected co-expression networks, one of them responsible for fast response to herbivory and anti-herbivory mechanisms and the other related to housekeeping genes, which present slower response to environmental variations. Integrating different levels of information such as gene expression patterns, gene assembly, transcriptomics, relationship among genes and phenotypes, functional relationships, among other information, enabled a wider visualization of S. frugiperda response to diet stimuli. The biological properties in the proposed networks are here described and discussed, as well as patterns of gene expression related to larval performance attributes.

高度综合性的害虫蛾Spodoptera frugiperda的基因表达数据集用于了解基因调控与幼虫寄主植物偏好的关系。的玉米和水稻株的转录数据草地夜蛾使用三种不同的基因网络推论方法分析了不同饮食饲养的幼虫，即共表达，加权共表达和贝叶斯网络，因为每种方法都提供了不同的数据可视化效果。使用这些方法，有可能确定两个松散互连的共表达网络，其中一个负责对草食和抗草食动物机制的快速响应，另一个与管家基因有关，后者对环境变化的响应较慢。整合不同水平的信息，例如基因表达模式，基因装配，转录组学，基因与表型之间的关系，功能关系以及其他信息，可以更广泛地可视化S. frugiperda对饮食刺激的反应。本文描述和讨论了拟议网络中的生物学特性，以及与幼虫性能属性有关的基因表达模式。