Background/Objectives: The larval stage plays a pivotal role in determining caste and sex in Apis mellifera. This study integrates RNA-seq, WGCNA, and alternative splicing analyses to explore gene expression differences among 2-day-old worker, drone, and queen larvae. Methods: RNA-seq was conducted on 2-day-old larvae from all three castes. Differential expression, WGCNA, and alternative splicing patterns were investigated. A deep learning TCN model was trained using WGCNA-derived modules and demonstrated high classification accuracy. Results: The TCN model highlighted a top-10 gene set, including PDHB, Fibroin3, and LOC724161. Significant caste- and sex-specific splicing events were detected in Tk, Csd, and Fem, with AF events being most prevalent. Splicing differences between sexes exceeded those observed among castes. Conclusions: The 2-day-old larval stage is crucial for both caste and sex differentiation in honeybees. This study identifies key genes and splicing events, offering new insights into the molecular mechanisms underlying caste formation and sex determination.

Background/Objectives: Early larval development is critical for caste and sex differentiation in honeybees. This study investigates molecular divergence in 4-day-old Apis mellifera larvae and introduces a customized deep learning model for hub-gene discovery. Methods: Genome-guided RNA-seq, DEGs, WGCNA, and splicing analyses were integrated. A hybrid convolution–attention model, ACmix-Swin, combined with WGAN-GP augmentation, was developed to classify larvae and prioritize caste-biased genes. Selected genes were validated by qPCR. Results: Significant caste- and sex-specific divergence was detected in cuticle formation, hormone metabolism, and reproductive signaling. ACmix-Swin achieved the highest accuracy among baseline models and consistently identified key regulators, including Vg, LOC725841, LOC412768, and LOC100576841. qPCR confirmed RNA-seq trends. Conclusions: Caste- and sex-specific transcriptional programs are established early in larval development. The ACmix-Swin framework provides an effective strategy for high-dimensional transcriptome interpretation and robust hub-gene identification.