Next Article in Journal
Phenotypic Characterization and Transcriptome Analysis of the Dwarf Mutant zmbrd1 in Maize
Previous Article in Journal
Heat Shock Protein 104 (Hsp104) in the Marine Diatom Ditylum brightwellii: Identification and Transcriptional Responses to Environmental Stress
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Genes
Volume 16
Issue 12
10.3390/genes16121409
genes-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Integrating WGCNA, TCN, and Alternative Splicing to Map Early Caste Programs in Day-2 Honeybee Larvae

by
Xiang Ding
1,
Jinyou Li
2,
Dan Yue
3 and
Runlang Su
3,*
1
University of Chinese Academy of Sciences, Beijing 100049, China
2
Faculty of Computing & Data Sciences, Boston University, Boston, MA 02215, USA
3
College of Animal Science, Yunnan Agricultural University, Kunming 650201, China
*
Author to whom correspondence should be addressed.
Genes 2025, 16(12), 1409; https://doi.org/10.3390/genes16121409
Submission received: 6 November 2025 / Revised: 21 November 2025 / Accepted: 25 November 2025 / Published: 26 November 2025
(This article belongs to the Section Epigenomics)
Browse Figures
Review Reports Versions Notes

Abstract

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.

Graphical Abstract

1. Introduction

Honeybees are highly social insects distinguished by a complex division of labor, with workers, queens, and males fulfilling distinct biological roles within the colony [1,2]. Although workers and queens are both Female, their caste fate is primarily determined by larval nutrition. Queens develop functional ovaries and are responsible for egg laying, whereas workers are sterile and perform tasks such as brood care, foraging, and nest construction. Males (drones) appear seasonally and mate with queens to sustain colony reproduction [3]. These caste differences are reflected at behavioral, physiological, and molecular levels.
Honeybee development progresses through the embryonic, larval, pupal, and adult stages. The larval stage is particularly dynamic, characterized by rapid growth and significant transcriptional and epigenetic shifts that shape caste differentiation. During the first three days, all larvae consume royal jelly, allowing Female larvae under three days to retain dual developmental potential [4]. The 2-day larval stage is considered a critical window for caste specification in both honeybees and bumblebees [5]. Although chromosomal conformation shows minimal differences between 2-day-old queen and worker larvae, chromatin architecture remains closely tied to transcriptional regulation [6]. DNA methylation also plays a central role, with queens generally showing lower methylation levels than workers [7]. At later larval stages, queens show widespread demethylation and a substantially greater number of differentially expressed genes than at the 2-day stage [8], underscoring the early larval period as a key period for regulatory divergence.
Previous transcriptomic studies have primarily focused on queen–worker comparisons, while information on male larvae remains limited. Thousands of genes differ between queen and worker larvae, influencing metabolism, development, and physiology. For example, antioxidant genes (MnSOD, CuZnSOD, catalase, Gst1) are associated with oxidative stress regulation [9], and mitochondrial translation factors such as AmIF-2mt and cytochrome c are linked to queen developmental rate [10]. Caste development consists of an initial dual-potential phase followed by irreversible commitment to a specific developmental trajectory [11]. In males, sex determination is governed by the complementary sex determination (Csd) gene, which regulates the downstream Feminizer (Fem) gene. Hemizygous Csd alleles trigger Female-pattern Fem splicing to produce males, while heterozygosity induces Female development [12,13]. Many sex-specific transcriptional differences arise from alternative splicing (AS) [14], and CRISPR/Cas9 knockdown of Fem has demonstrated its role in regulating the splicing of multiple downstream genes involved in sexual differentiation [15].
Alternative splicing is a key post-transcriptional mechanism that greatly expands proteome complexity. Approximately 95% of human genes undergo AS [16], and AS contributes to development and diverse cellular processes by generating multiple mRNA isoforms from a single gene [17]. Splicing regulation is essential across species [18]. In honeybees, AS is associated with epigenetic modifications, such as DNA and m6A methylation, that shape caste-specific transcript variation [7,19,20].
To elucidate transcriptional and splicing differences across castes, this study analyzes transcriptomes of 2-day-old worker, queen, and male larvae of the western honeybee using WGCNA and alternative splicing approaches. This study aims to identify differentially expressed genes, characterize sex-related transcriptional signatures, and investigate how AS contributes to caste-specific developmental pathways.

2. Materials and Methods

2.1. Sample Collection

Transcriptome data from 2-day-old Italian honeybee larvae were retrieved from the NCBI database, with six biological replicates for each caste (workers: SRS1249139; queens: SRS1263242; drones: SRS1263244) [21]. To validate the RNA-seq findings, we also collected an independent qPCR cohort consisting of 2-day-old Apis mellifera worker, queen, and drone larvae, each group including six biological replicates, under matched environmental and experimental conditions.
For sex-level analyses, workers and queens, both genetically Female (diploid), were combined into a non-drone (ND) group to increase Female sample size and better capture sex-related expression patterns shared across Female larvae. In comparison, caste-level comparisons (queen vs. worker) were performed without pooling to preserve caste-specific transcriptional features and avoid confounding effects between sex and caste.
Larvae were sampled from Apis mellifera colonies maintained near Yiwu Industrial and Commercial College. Following collection, samples were immediately processed for RNA extraction and qPCR under controlled laboratory conditions. Rearing conditions were maintained at approximately 34 °C and 70% relative humidity, with no additional environmental enrichment. All sample processing and qPCR assays were carried out by Anhui Gaohe Biotechnology Co., Ltd. (Anqing, China).

2.2. Mapping and Quality Control of RNA-Seq Data

Quality control of the clean sequencing data was performed using FastQC to generate base composition and quality distribution plots. Raw reads were processed with Fastp (v0.23.2) for adapter removal and quality trimming using the following parameters: automatic adapter detection for paired-end reads, Q20 base trimming, filtering of reads with >40% low-quality bases or >5 ambiguous bases, and retention of reads ≥50 bp. Clean reads were aligned to the Apis mellifera reference genome (Amel_HAv3.1) using Hisat2 (v2.2.1) [22]. By default, Hisat2 uses a 20 bp seed length, permits soft clipping, applies mismatch penalties of 6 and 2, and reports up to 5 valid alignments per read. Alignment outputs in SAM format were converted to BAM files, sorted with Samtools (v1.6) [23], and quantified with FeatureCounts [24] to obtain gene-level read counts. Differential expression analysis was performed using the DESeq2 algorithm integrated within Trinity (v2.15.1) [25,26]. p-values were adjusted using the Benjamini–Hochberg false discovery rate (FDR), and genes with FDR < 0.05 and |log2FC| > 1 were considered significantly differentially expressed. To statistically compare the relative contributions of sex and caste to transcriptional variation, a two-way ANOVA was performed for each gene with sex (male vs. female) and caste (worker vs. queen vs. drone) as factors. Effect sizes were quantified using partial eta-squared (ηp2), and the significance of main effects was assessed at α = 0.05.

2.3. PCA

To assess sample consistency, principal component analysis (PCA) was conducted on the gene expression matrix. Before performing PCA, the data were standardized using the scale function in R to ensure comparability across all dimensions.

2.4. GO Enrichment Analysis of Differentially Expressed and Alternatively Spliced Genes

The longest CDS-derived protein sequences from the Apis mellifera reference genome (Amel_HAv3.1) were annotated for Gene Ontology (GO) using emapper.py (v2.1.10) [27] with the EggNOG database [28]. The resulting annotation file was converted into org.db format, and GO enrichment analyses were performed using the clusterProfiler package (v4.4) in R. For both GO and KEGG enrichment analyses, multiple testing correction was applied using the Benjamini–Hochberg FDR method, and only terms with FDR < 0.05 (p-value < 0.05 and adjusted p-value < 0.05) were retained.

2.5. Alternative Splicing Analysis

Alternative splicing analysis was conducted using SUPPA (v2.4) with the reference genome annotation (GTF) [29,30]. Major SUPPA splicing event types, including skipped exon (SE), mutually exclusive exon (MXE), alternative 5′ splice site (A5), alternative 3′ splice site (A3), and retained intron (RI), were quantified. Transcript abundance was estimated using Salmon (v1.10.1), and ΔPSI values were calculated for all events. From individual biological replicates (n = 6 per phenotype group: worker, drone, queen). Percent-spliced-in (PSI) values were calculated for each splicing event in each replicate, and group-level ΔPSI values were derived by comparing mean PSI values between phenotype pairs (worker vs. queen, worker vs. drone, drone vs. queen). Differential splicing was identified using a ΔPSI threshold of ≥0.1, together with statistical significance (p < 0.05) and Benjamini–Hochberg FDR correction. Only events meeting both ΔPSI and FDR criteria were considered significant.

2.6. Weighted Gene Co-Expression Network Analysis (WGCNA) and TCN Modeling

After removing outlier samples through pairwise correlation analysis and hierarchical clustering, weighted co-expression networks were constructed using the WGCNA R package(v1.73) (https://cran.r-project.org/web/packages/WGCNA/index.html) (accessed on 23 November 2025) (networkType = “signed”, mergeCutHeight = 0.25, minModuleSize = 30, deepSplit = 2). Automatic blockwise module detection was applied, using a data-driven soft threshold to approximate a scale-free topology. Module eigengenes (MEs) were calculated and correlated with experimental conditions to identify trait-associated modules. Gene–gene regulatory networks from selected modules were exported and visualized in Cytoscape (v3.10.1).
To refine biomarkers from trait-associated modules and develop a phenotype classifier, a Temporal Convolutional Network (TCN) was trained using the module gene expression matrices (https://github.com/surunlang-creator/Drone_torch) (accessed on 23 November 2025). Expression values were log-transformed and z-scored for each gene, and a fixed random seed (5678) ensured reproducibility. Data were partitioned into training, validation, and test sets at a 6:2:2 ratio with phenotype stratification (worker, drone, queen) to maintain balanced class representation across splits. The TCN architecture consisted of three dilated causal convolutional layers with kernel size 3, dilation rates of 1, 2, and 4 to capture multi-scale temporal dependencies, and channel dimensions of 64, 64, and 128, respectively. Each convolutional layer was followed by batch normalization and dropout (rate = 0.3) for regularization, with final classification performed through adaptive average pooling and a fully connected layer. The model was implemented in PyTorch (2.9.1), optimized using the Adam optimizer (learning rate = 0.001, weight decay = 1 × 10−4) with ReduceLROnPlateau scheduler (factor = 0.5, patience = 10), and trained for up to 100 epochs with early stopping (patience = 20) based on validation loss to prevent overfitting. Model performance was assessed on the held-out test set using accuracy metrics, precision, recall, F1-score, training/validation loss curves, and a confusion matrix. Gene-level importance scores were generated from the trained model by calculating composite scores integrating prediction accuracy, expression variance (R2 values), and pattern-specific weights, with the top 10 features per phenotype retained for downstream interpretation and network visualization. These candidate genes were subsequently used to guide qPCR validation experiments.

3. Results

3.1. Overall RNA-Seq Results and Quality Control

Illumina RNA sequencing produced an average of 3.42 Gbp of raw reads per sample. After quality filtering, approximately 3.38 Gbp of clean reads per sample were retained for downstream transcriptome analysis. These clean reads demonstrated high sequencing quality, with average Q20 and Q30 values both reaching 100%, and an average GC content of 38.00% (Supplemental Table S1). Alignment to the reference genome with HISAT2 showed that more than 88% of transcripts were uniquely mapped, as determined by sliding-window density analysis. Across all 18 samples, 645 million clean reads were uniquely aligned, yielding an average mapping rate of 93% (Supplemental Table S2).

3.2. Hierarchical Clustering and Principal Component Analysis of Samples

Hierarchical clustering and PCA (Figure 1) demonstrated strong within-group consistency and clear separation among queen, worker, and drone samples. PCA was performed on all expressed genes following standard scaling, with PC1 and PC2 accounting for the majority of the variance. Outlier detection based on sample-to-sample distance matrices and projections onto the first two principal components revealed no abnormal samples. The tight clustering of biological replicates within each caste further indicated the absence of batch effects. These findings establish a reliable basis for subsequent differential expression and alternative splicing analyses.

3.3. Analysis of Differentially Expressed Genes (DEGs)

To investigate gene expression differences among workers, queens, and drones, a differential expression analysis was performed, and DEGs across three pairwise comparisons, drones vs. queens, workers vs. drones, and workers vs. queens (Figure 2A), were compared. The smallest expression difference was observed between workers and queens, with 247 DEGs identified (DESeq2, adjusted p < 0.05, |log2FC| > 1), whereas drones and queens showed the largest divergence, with 782 DEGs detected. This difference was statistically significant (x2 test, p < 0.001), indicating that sex-based differences exceed caste-based differences by 3.2-fold. Volcano plot analysis further highlighted several genes demonstrating substantial fold changes (Figure 2B). Genes with high fold change between drones and queens included LOC551527, LOC410733, LOC411233, and LOC411188. Genes strongly differentiated between workers and drones included CPR5, LOC100578919, LOC411884, LOC726134, and Fem, while LOC724536, LOC100578046, and vg showed prominent differences between workers and queens. These highly variable genes are likely to play important roles in honeybee sex differentiation, physiological traits, and behavioral regulation.

3.4. Gene Expression GO Analysis

To explore the functional roles of the differentially expressed genes among the three castes, the top 11 most highly expressed genes were selected for GO enrichment analysis (Figure 3). The DEGs between drones and queens were predominantly enriched in biological processes related to development and growth, sex differentiation, male sex differentiation, primary sexual characteristic development, and male sex determination. DEGs identified between workers and drones were enriched in pathways associated with sex differentiation, male sex differentiation, primary male sexual characteristic development, male anatomical structure morphogenesis, and primary sex determination. In comparison, DEGs between workers and queens were mainly enriched in processes linked to embryonic organ morphogenesis, brain development, and vitellogenesis.

3.5. WGCNA

To investigate genetic correlations and caste-specific expression patterns among workers, drones, and queens, WGCNA was performed on all aligned genes (9,939 in total). After sample clustering (Figure 4A), missing values were processed and outliers removed. To ensure that the network approximated a scale-free topology, a soft-threshold power of 24 was selected (scale-free R2 = 0.80) (Figure 4B,C). Using the one-step construction method, a co-expression matrix was generated, and 9 gene modules were identified via dynamic tree-cutting (Figure 4A,D). The turquoise module (1751 genes) and gray module (5412 genes) showed the strongest association with drones (correlation = 0.82, p = 3 × 10−5) (Figure 4D). The brown module (368 genes) displayed the highest correlation with queens (correlation = 0.82, p = 3 × 10−5), while the yellow module (210 genes) was most strongly associated with workers (correlation = 0.70, p = 0.001). These module–trait relationships highlight significant specificity in gene expression across castes, suggesting that unique transcriptional programs underpin the distinct biological functions and behaviors of each group.
A TCN deep learning model was applied to further evaluate the phenotype-associated modules identified by WGCNA. The dataset was divided into training, validation, and test sets at a 9:1:1 ratio. As shown in Figure 5A,B, both training and validation loss rapidly converged toward zero without evidence of overfitting, and accuracy for both sets reached nearly 100% within the first 30 epochs. The model achieved perfect classification accuracy (100%) on the independent test set, demonstrating strong generalization performance. Using the trained TCN model, top-ranked representative genes were extracted for the worker, drone, and queen phenotypes. These indicator genes were visualized using neural network diagrams (Supplemental Figure S1), and their expression differences among castes were statistically significant (Supplemental Figures S2 and S3). qPCR validation (Figure 6; Supplemental Figures S2 and S3; Supplemental Table S3) further confirmed the accuracy and reliability of the TCN-derived gene candidates.

3.6. Statistical Analysis of Differential Genes and Events in Alternative Splicing Events

To investigate differences in gene isoform expression among workers, drones, and queens, a detailed analysis of alternative splicing patterns was performed. Consistent with the differential expression results, workers and queens showed the smallest divergence, with the fewest alternatively spliced genes and corresponding transcripts, 25 and 20, respectively (Figure 7A). In comparison, drones and workers showed the greatest differences, with 51 alternatively spliced genes and 46 associated transcripts (Figure 7A). Among all splicing categories, AF (Alternative First Exon) events were the most prevalent across castes, totaling 27 in drones, 20 in queens, and 15 in workers (Figure 7B). These patterns likely reflect caste-specific regulatory requirements shaped by distinct physiological roles. In particular, the higher number of alternatively spliced genes and transcripts, especially the increased AF events, between drones and workers may indicate the need for more complex regulatory mechanisms governing drone reproduction, reproductive organ development, and mating-associated behaviors.

3.7. GO Enrichment Analysis of Alternative Splicing

GO enrichment analysis was performed to explore the functional implications of alternative splicing events among the three honeybee castes. The results revealed both shared and caste-specific patterns related to developmental and functional differentiation. Drones and queens displayed strong similarities in reproductive and developmental regulatory pathways, particularly those associated with reproductive organ growth and regulation. Drones and workers showed overlap in processes linked to developmental maturation and core reproductive functions, reflecting common features of sexual differentiation. Queens and workers were closely aligned in pathways related to structural formation and functional specialization, consistent with their roles in morphogenesis and social division of labor (Figure 8). These results provide important molecular insights into the mechanisms driving social differentiation and functional regulation in honeybees.

3.8. Alternative Splicing of Genes Related to Sex Determination

To investigate how alternative splicing contributes to sex determination and differentiation among the three honeybee castes, three genes—two known sex-determination genes (Csd and Fem) and an additional gene, Tk (Figure 9A,B)—were analyzed. All three genes demonstrated alternative splicing events across castes. The Csd gene (chr3:11,135,000–11,146,004) showed a unique intron retention event in drones and an exclusive exon-skipping event in queens. Furthermore, at the genomic region chr3:11,781,139–11,780,712, queens lacked an intron-retention event present in both workers and drones. For Fem, drones displayed the greatest number of exon-skipping and intron-retention events.
In comparison, workers and queens shared similar splicing patterns, with workers demonstrating one additional exon-skipping event. In the Tk gene (chr7:3,735,230–3,735,382 and chr7:3,735,570–3,735,914), queens lacked two intron retention events observed in workers. These results highlight caste-specific splicing patterns among genes with large fold changes and suggest their involvement in honeybee developmental and caste-differentiation processes.
To validate these predicted splicing events, primers targeting specific transcript regions were designed, and qPCR analysis (Figure 10; Supplemental Table S4) was performed. The qPCR results confirmed the predicted alternative splicing patterns, showing complete consistency with the caste-specific events identified for all three genes.

4. Discussion

To identify genes with substantial fold changes and caste-specific alternative splicing patterns in 2-day-old western honeybee workers, drones, and queens, transcriptomic datasets from the NCBI database were first retrieved. More stringent data-processing standards were applied than those used in previous studies. This approach yielded a larger set of differentially expressed genes compared with previous reports [21] (Figure 2A). Further, PCA revealed a clear intrinsic structure among samples, with strong within-group consistency for workers, queens, and drones. These results confirm both the robustness of the data and the reliability of the filtering strategies used in this study.
In the analysis of differentially expressed genes (Figure 2A), more stringent criteria were applied to the raw data, resulting in a substantial increase in the number of detected DEGs compared with the findings of He et al. The number of DEGs between males and queens increased by 64%, while the numbers for workers vs. males and workers vs. queens nearly doubled. Among the three castes, males and queens demonstrated the highest number of DEGs, whereas workers and queens showed the fewest. In comparison to the observations by Christina Vleurinck et al. [31], reporting that caste-related DEGs during the pupal stage were shared mainly between sexes, these results demonstrate that at the 2-day larval stage, sex-related differences significantly exceed caste-related differences (two-way ANOVA across all expressed genes: mean sex effect ηp2 = 0.58 ± 0.12, mean caste effect ηp2 = 0.31 ± 0.09; paired t-test, p < 0.001). Specifically, the drone vs. queen comparison yielded 782 DEGs, significantly more than the 247 DEGs between workers and queens (Fisher’s exact test, p = 3.4 × 10−8, odds ratio = 3.17). In the volcano plots (Figure 2B), several high-fold change genes associated with honeybee growth, development, and sex determination were identified. Among them, vg encodes vitellogenin, a major Female-specific yolk glycoprotein precursor. As a Female-specific glycolipoprotein, vitellogenin supports fat body development and enhances immunity and stress resistance in honeybees [32,33]. Another essential gene is Fem, a key regulator of sex determination [34]. Its expression is controlled by the complementary sex determiner (Csd) gene, and only Female-specific splicing of Fem produces functional Fem protein required for Female development [34]. CPR5 encodes an epidermal protein that interacts with chitin and contributes to exoskeleton formation and stability, playing important roles in structural integrity, water retention, and environmental protection [35]. LOC411233 encodes MPP10, a core component of the U3 small nucleolar ribonucleoprotein complex involved in ribosomal RNA processing and ribosome assembly [36]. Since ribosomes are essential for protein synthesis, this gene is critical for normal cell growth and development. Several highly expressed genes with previously unknown functions, such as LOC724536 and LOC725439, were also identified that may contribute to developmental processes in honeybees.
GO enrichment of DEGs among workers, drones, and queens (Figure 3) showed that most of the differences between drones and queens, and between drones and workers, involved sex differentiation, indicating pronounced sex-specific divergence during the larval stage. In comparison, DEGs between workers and queens were primarily enriched in pathways related to brain development and organogenesis. These differences likely reflect nutritional divergence: the protein and 10-HDA content in the diet of 2-day-old queen larvae is significantly higher than that of worker larvae [37], driving caste-specific developmental outcomes. This nutritional influence complies with the worker–queen DEG patterns observed in the volcano plot (Figure 2B). These findings indicate that both sex differentiation and caste-specific gene expression differences are already well established by the 2-day larval stage in workers, drones, and queens.
Integrating WGCNA with deep learning enabled us to substantially narrow the candidate gene pool while maintaining strong phenotype-discriminative power, a strategy that has proven effective in various disease-gene discovery studies [38,39,40,41]. Using phenotype-associated modules derived from WGCNA, the TCN classifier identified a concise set of high-priority genes. Investigation of the top 10 weighted features revealed several biologically meaningful signals. PDHB showed the most striking differential expression pattern, with expression levels in non-queen larvae 6.87-fold higher than in queen larvae (log2FC = 2.78, adjusted p-value = 1.2 × 10−5, 95% CI: 5.93–7.81-fold). PDHB encodes the β-subunit of the pyruvate dehydrogenase complex, a central metabolic enzyme that channels glycolytic carbon into the TCA cycle and interfaces with lipid and energy metabolism. Its reduced expression in 2-day-old queen larvae aligns with previous findings that queens and workers adopt distinct metabolic strategies early in development [42]. Studies in other biological systems also link PDH/PDHB activity to developmental processes, including skeletal and tissue growth, suggesting that early PDHB modulation may couple nutrient signaling to downstream morphogenetic pathways [42]. Although direct causal relationships in honeybees remain unconfirmed, these results demonstrate that caste-specific metabolic–developmental divergence is already established at the larval stage.
Caste-biased expression of Fibroin3, a member of the fibroin family that forms the primary structural component of the insect silk gland and contributes to the architecture of the cocoon and nest, was also identified. In honeybees, larval silk plays an essential role in reinforcing comb stability; therefore, early divergence in Fibroin3 expression provides molecular support for caste-specific differences in structural and behavioral potential related to nest building [43,44]. This interpretation is in good agreement with qPCR results, which showed elevated Fibroin3 expression in the non-drone group. This pattern aligns with the adult division of labor, in which drones lack functional wax glands and do not participate in comb construction. In comparison, workers develop well-developed wax glands, which are essential for nest building. However, direct evidence linking fibroin genes to wax gland ontogeny remains limited. A more cautious interpretation is that Fibroin3 marks early specification of structural and behavioral programs associated with nest architecture rather than directly influencing wax gland formation [45].
These results demonstrate how a WGCNA-to-deep learning framework can distill high-dimensional transcriptomic data into a compact, interpretable, and phenotype-predictive gene set, while highlighting mechanistic hypotheses linking metabolic regulation (PDHB) to caste-specific structural programs (Fibroin3). Future studies should integrate multi-cohort validation with functional perturbation approaches, such as RNAi, CRISPR, or nutrient and hormone manipulations, to test causality, and employ cell-type-resolved analyses to map these molecular signatures onto the developmental lineages that drive early caste divergence.
In the analysis of differential alternative splicing events (Figure 7A), the fewest selectively spliced genes and transcripts between workers and queens were observed. In comparison, males showed substantially more selectively spliced genes than either workers or queens. This pattern mirrors the differential gene expression results (Figure 2A), indicating that during the 2-day larval stage, sex-specific differences in alternative splicing are more pronounced than caste-specific differences. Alternative First Exon (AF) events were the most frequent among all splicing types (Figure 5B). Because 2-day-old honeybee larvae are undergoing rapid growth and developmental transitions, selective splicing of the first exon can generate protein isoforms with distinct functional domains or subcellular localizations, increasing protein diversity [46]. GO enrichment analysis of these alternative splicing events further revealed that the associated genes participate in a wide range of processes, including growth, development, and sexual reproduction. These findings suggest that alternative splicing contributes broadly to the regulation of sex differentiation and other physiological changes in honeybee larvae.
Because Fem showed substantial expression differences between workers and males in Figure 2B and Figure 9B, alternative splicing patterns in Fem, its upstream regulator Csd, and the behavior-associated gene Tk were further analyzed. Csd is a honeybee-specific sex-determination gene [44], and its RS domain and proline-rich region are known to mediate protein-binding interactions [45,46]. Variations in these amino acid sequences may therefore alter protein–protein interactions that are critical for splicing regulation. In this study, both Csd and Fem demonstrated significantly higher expression in males than in queens and workers (Figure 7), while Tk showed no substantial differences in overall expression. Consistent with this, the alternative splicing patterns of the three genes showed a similar trend (Figure 9C): Tk had relatively simple splicing variants across castes, whereas Csd and Fem displayed much more complex splicing profiles, particularly in males. These findings suggest a strong relationship between gene expression levels and the splicing complexity of their transcripts.
All three genes, Csd, Fem, and Tk, demonstrated caste-specific splicing variants, with males showing the most complex patterns. This supports the idea that alternative splicing influences not only caste development [11] but also sex differentiation. A mechanistic association between the alternative splicing variants of Fem and Csd is further hypothesized. Previous studies have shown that the binary switch in honeybee sex determination is controlled through selective splicing of Fem transcripts in response to Csd heterozygosity [45,46,47,48,49,50]. Similarly, in Drosophila, the tra gene, homologous to Csd, acts as a switch gene required for Female development through sex-specific splicing regulation [51,52]. Our analysis demonstrates that Csd and Fem undergo distinct alternative splicing patterns in workers, males, and queens, underscoring the utility of splicing-based approaches for studying sex determination in honeybees. These findings also provide new avenues for future research on developmental regulation and behavioral differentiation in honeybee castes.
In summary, the combined results from WGCNA and alternative splicing analyses demonstrate that 2-day-old worker, drone, and queen larvae are at a pivotal stage of sex differentiation, physiological transition, and organ development. Both the overall number of differentially expressed genes and the number of differentially spliced genes indicate that sex-related differences are more pronounced than caste-related differences at this early stage. Tissue- and organ-specific developmental pathways appear to be shaped by distinct patterns of selective splicing, which likely drive the divergent gene expression profiles observed among the three castes. These findings improve our understanding of the molecular mechanisms underlying early larval development in workers, drones, and queens, and provide valuable insights into how alternative splicing influences developmental trajectories in honeybee larvae.

5. Conclusions

In this study, transcriptome data from 2-day-old worker, drone, and queen larvae of the western honeybee were reanalyzed using enhanced filtering, WGCNA, and alternative splicing approaches. Both caste- and sex-specific differences in gene expression and splicing, including key regulatory genes such as vg, Fem, and Csd, were identified. These findings underscore the substantial transcriptional and post-transcriptional regulation occurring at this early larval stage and provide valuable insights into the molecular mechanisms governing honeybee caste differentiation and sex determination.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/genes16121409/s1. Figure S1: Gene Interaction Network of TCN across Honey Bee Castes; Figure S2: Comprehensive Gene Expression Profiles of the Top 10 Caste-Specific Genes; Figure S3: qPCR Validation of Caste-Specific Gene Expression in Honey Bees following TCN Treatment; Table S1: Summary of Sequencing Data Quality Metrics; Table S2: Summary of Sequencing Read Alignment Rates; Table S3: qPCR Validation of Top 10 Genes Following TCN Treatment in Honey Bee Castes; Table S4: qPCR Validation of Alternatively Spliced Isoforms across Honey Bee Castes.

Author Contributions

X.D.: Data curation, Visualization, and Fornal analysis. J.L.: Visualization; Writing—review and editing. D.Y.: Writing—original draft. R.S.: Writing—original draft, Writing—review and editing. All authors have read and agreed to the published version of the manuscript.

Funding

This research is supported by the Yiwu City Science and Technology Bureau and Yiwu Industrial and Commercial College Program under Grant Nos. XM2025DCZ0227 and XJKJ2502YB. This work was supported by Jinhua Key Laboratory of Robot Intelligent Welding Technology and by Jinhua Public Welfare Technology Application Research Project in 2024.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Raw data have been deposited to National Center for Biotechnology Information (NCBI) under the BioProject numbers SRS1249139, SRS1263242, SRS1263244, SRS1263211, SRS1263243 and SRS1263256.

Conflicts of Interest

The authors declare that they have no competing interests.

References

  1. Brouwers, E.V.M.; Ebert, R.; Beetsma, J. Behavioural and physiological aspects of nurse bees in relation to the composition of larval food during caste differentiation in the honeybee. J. Apic. Res. 1987, 26, 11–23. [Google Scholar] [CrossRef]
  2. Chen, X.; Hu, Y.; Zheng, H.Q.; Cao, L.F.; Niu, D.F.; Yu, D.L.; Sun, Y.Q.; Hu, S.N.; Hu, F.L. Transcriptome comparison between honey bee queen-and worker-destined larvae. Insect Biochem. Mol. Biol. 2012, 42, 665–673. [Google Scholar] [CrossRef]
  3. Rangel, J.; Fisher, A. Factors affecting the reproductive health of honey bee (Apis mellifera) drones—A review. Apidologie 2019, 50, 759–778. [Google Scholar] [CrossRef]
  4. He, X.J.; Zhou, L.B.; Pan, Q.Z.; Barron, A.B.; Yan, W.Y.; Zeng, Z.J. Making a queen: An epigenetic analysis of the robustness of the honeybee (Apis mellifera) queen developmental pathway. Mol. Ecol. 2017, 26, 1598–1607. [Google Scholar] [CrossRef]
  5. Cnaani, J.; Robinson, G.E.; Hefetz, A. The critical period for caste determination in Bombus terrestris and its juvenile hormone correlates. J. Comp. Physiol. A 2000, 186, 1089–1094. [Google Scholar] [CrossRef] [PubMed]
  6. Zhang, Y.; He, X.J.; Barron, A.B.; Li, Z.; Jin, M.J.; Wang, Z.L.; Huang, Q.; Zhang, L.Z.; Wu, X.B.; Yan, W.Y.; et al. The diverging epigenomic landscapes of honeybee queens and workers revealed by multiomic sequencing. Insect Biochem. Mol. Biol. 2023, 155, 103929. [Google Scholar] [CrossRef] [PubMed]
  7. Wang, M.; Xiao, Y.; Li, Y.; Wang, X.; Qi, S.; Wang, Y.; Zhao, L.; Wang, K.; Peng, W.; Luo, G.-Z.; et al. RNA m6A modification functions in larval development and caste differentiation in honeybee (Apis mellifera). Cell Rep. 2021, 34, 108580. [Google Scholar] [CrossRef]
  8. Shi, Y.Y.; Yan, W.Y.; Huang, Z.Y.; Wang, Z.L.; Wu, X.B.; Zeng, Z.J. Genomewide analysis indicates that queen larvae have lower methylation levels in the honey bee (Apis mellifera). Sci. Nat. 2013, 100, 193–197. [Google Scholar] [CrossRef]
  9. Santos, D.E.; de Oliveira Souza, A.; Tibério, G.J.; Alberici, L.C.; Hartfelder, K. Differential expression of antioxidant system genes in honey bee (Apis mellifera L.) caste development mitigates ROS-mediated oxidative damage in queen larvae. Genet. Mol. Biol. 2020, 43, e20200173. [Google Scholar] [CrossRef] [PubMed]
  10. Corona, M.; Estrada, E.; Zurita, M. Differential expression of mitochondrial genes between queens and workers during caste determination in the honeybee Apis mellifera. J. Exp. Biol. 1999, 202, 929–938. [Google Scholar] [CrossRef]
  11. Cameron, R.C.; Duncan, E.J.; Dearden, P.K. Biased gene expression in early honeybee larval development. BMC Genom. 2013, 14, 903. [Google Scholar] [CrossRef] [PubMed]
  12. El Agoze, M.; Drezen, J.M.; Renault, S.; Periquet, G. Analysis of the reproductive potential of diploid males in the wasp Diadromus pulchellus (Hymenoptera: Ichneumonidae). Bull. Entomol. Res. 1994, 84, 213–218. [Google Scholar] [CrossRef]
  13. Suzuki, Y.; Yamada, T.; Suzuki, M.G. In vitro comparison of sex-specific splicing efficiencies of Fem. pre-mRNA under monoallelic and heteroallelic conditions of Csd, a master sex-determining gene in the honeybee. J. Dev. Biol. 2023, 11, 10. [Google Scholar] [CrossRef] [PubMed]
  14. Netschitailo, O.; Raub, S.; Kaftanoglu, O.; Beye, M. Sexual diversification of splicing regulation during embryonic development in honeybees (Apis mellifera), A haplodiploid system. Insect Mol. Biol. 2022, 31, 170–176. [Google Scholar] [CrossRef]
  15. Cheng, F.-P.; Hu, X.-F.; Pan, L.-X.; Gong, Z.-X.; Qin, K.-X.; Li, Z.; Wang, Z.-L. Transcriptome changes of Apis mellifera female embryos with Fem. gene knockout by CRISPR/Cas9. Int. J. Biol. Macromol. 2023, 229, 260–267. [Google Scholar] [CrossRef]
  16. Zhu, G.-Q.; Zhou, Y.-J.; Qiu, L.-X.; Wang, B.; Yang, Y.; Liao, W.-T.; Luo, Y.-H.; Shi, Y.-H.; Zhou, J.; Fan, J.; et al. Prognostic alternative mRNA splicing signature in hepatocellular carcinoma: A study based on large-scale sequencing data. Carcinogenesis 2019, 40, 1077–1085. [Google Scholar] [CrossRef]
  17. Wright, C.J.; Smith, C.W.J.; Jiggins, C.D. Alternative splicing as a source of phenotypic diversity. Nat. Rev. Genet. 2022, 23, 697–710. [Google Scholar] [CrossRef]
  18. Kalsotra, A.; Cooper, T.A. Functional consequences of developmentally regulated alternative splicing. Nat. Rev. Genet. 2011, 12, 715–729. [Google Scholar] [CrossRef]
  19. Foret, S.; Kucharski, R.; Pellegrini, M.; Feng, S.H.; Jacobsen, S.E.; Robinson, G.E.; Maleszka, R. DNA methylation dynamics, metabolic fluxes, gene splicing, and alternative phenotypes in honey bees. Proc. Nati. Acad. Sci. USA 2012, 109, 4968–4973. [Google Scholar] [CrossRef]
  20. Kucharski, R.; Maleszka, J.; Foret, S.; Maleszka, R. Nutritional control of reproductive status in honeybees via DNA methylation. Science 2008, 319, 1827–1830. [Google Scholar] [CrossRef]
  21. He, X.J.; Jiang, W.J.; Zhou, M.; Barron, A.B.; Zeng, Z.J. A comparison of honeybee (Apis mellifera) queen, worker and drone larvae by RNA-Seq. Insect Sci. 2019, 26, 499–509. [Google Scholar] [CrossRef] [PubMed]
  22. Kim, D.; Langmead, B.; Salzberg, S.L. HISAT: A fast spliced aligner with low memory requirements. Nat. Methods 2015, 12, 357–360. [Google Scholar] [CrossRef] [PubMed]
  23. Li, H.; Handsaker, B.; Wysoker, A.; Fennell, T.; Ruan, J.; Homer, N.; Marth, G.; Abecasis, G.; Durbin, R.; 1000 Genome Project Data Processing Subgroup. The sequence alignment/map format and SAMtools. Bioinformatics 2009, 25, 2078–2079. [Google Scholar] [CrossRef] [PubMed]
  24. Liao, Y.; Smyth, G.K.; Shi, W. FeatureCounts: An efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 2014, 30, 923–930. [Google Scholar] [CrossRef]
  25. Love, M.I.; Huber, W.; Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014, 15, 550. [Google Scholar] [CrossRef]
  26. Haas, B.J.; Papanicolaou, A.; Yassour, M.; Grabherr, M.; Blood, P.D.; Bowden, J.; Couger, M.B.; Eccles, D.; Li, B.; Lieber, M.; et al. De novo transcript sequence reconstruction from RNA-seq using the Trinity platform for reference generation and analysis. Nat. Protoc. 2013, 8, 1494–1512. [Google Scholar] [CrossRef]
  27. Cantalapiedra, C.P.; Hernández-Plaza, A.; Letunic, I.; Bork, P.; Huerta-Cepas, J. EggNOG-mapper v2: Functional annotation, orthology assignments, and domain prediction at the metagenomic scale. Mol. Biol. Evol. 2021, 38, 5825–5829. [Google Scholar] [CrossRef]
  28. Huerta-Cepas, J.; Szklarczyk, D.; Heller, D.; Hernández-Plaza, A.; Forslund, S.K.; Cook, H.; Mende, D.R.; Letunic, I.; Rattei, T.; Jensen, L.J.; et al. eggNOG 5.0: A hierarchical, functionally and phylogenetically annotated orthology resource based on 5090 organisms and 2502 viruses. Nucleic Acids Res. 2019, 47, D309–D314. [Google Scholar] [CrossRef]
  29. Trincado, J.L.; Entizne, J.C.; Hysenaj, G.; Singh, B.; Skalic, M.; Elliott, D.J.; Eyras, E. SUPPA2: Fast, accurate, and uncertainty-aware differential splicing analysis across multiple conditions. Genome Biol. 2018, 19, 40. [Google Scholar] [CrossRef]
  30. Alamancos, G.P.; Pagès, A.; Trincado, J.L.; Bellora, N.; Eyras, E. Leveraging transcript quantification for fast computation of alternative splicing profiles. RNA 2015, 21, 1521–1531. [Google Scholar] [CrossRef]
  31. Vleurinck, C.; Raub, S.; Sturgill, D.; Oliver, B.; Beye, M. Linking genes and brain development of honeybee workers: A whole-transcriptome approach. PLoS ONE 2016, 11, e0157980. [Google Scholar] [CrossRef] [PubMed]
  32. Nunes, F.M.F.; Ihle, K.E.; Mutti, N.S.; Simões, Z.L.P.; Amdam, G.V. The gene vitellogenin affects microRNA regulation in honey bee (Apis mellifera) fat body and brain. J. Exp. Biol. 2013, 216, 3724–3732. [Google Scholar] [CrossRef]
  33. Seehuus, S.-C.; Norberg, K.; Gimsa, U.; Krekling, T.; Amdam, G.V. Reproductive protein protects functionally sterile honey bee workers from oxidative stress. Proc. Natl. Acad. Sci. USA 2006, 103, 962–967. [Google Scholar] [CrossRef] [PubMed]
  34. Gempe, T.; Hasselmann, M.; Schiøtt, M.; Hause, G.; Otte, M.; Beye, M. Sex determination in honeybees: Two separate mechanisms induce and maintain the female pathway. PLoS Biol. 2009, 7, e1000222. [Google Scholar] [CrossRef] [PubMed]
  35. Zhao, X.M.; Gou, X.; Qin, Z.Y.; Li, D.Q.; Wang, Y.; Ma, E.B.; Li, S.; Zhang, J.Z. Identification and expression of cuticular protein genes based on Locusta migratoria transcriptome. Sci. Rep. 2017, 7, 45462. [Google Scholar] [CrossRef]
  36. Hunziker, M.; Barandun, J.; Petfalski, E.; Tan, D.Y.; Delan-Forino, C.; Molloy, K.R.; Kim, K.H.; Dunn-Davies, H.; Shi, Y.; Chaker-Margot, M.; et al. UtpA and UtpB chaperone nascent pre-ribosomal RNA and U3 snoRNA to initiate eukaryotic ribosome assembly. Nat. Commun. 2016, 7, 12090. [Google Scholar] [CrossRef]
  37. Wang, Y.; Ma, L.T.; Zhang, W.X.; Cui, X.P.; Wang, H.F.; Xu, B.H. Comparison of the nutrient composition of royal jelly and worker jelly of honey bees (Apis mellifera). Apidologie 2016, 47, 48–56. [Google Scholar] [CrossRef]
  38. Lv, J.-H.; Hou, A.-J.; Zhang, S.-H.; Dong, J.-J.; Kuang, H.-X.; Yang, L.; Jiang, H. WGCNA combined with machine learning to find potential biomarkers of liver cancer. Medicine 2023, 102, e36536. [Google Scholar] [CrossRef]
  39. Wang, Y.X.; Zhu, H.X.; Ren, J.L.; Ren, M.H. Integrative machine learning models predict prostate cancer diagnosis and biochemical recurrence risk: Advancing precision oncology. NPJ Digit. Med. 2025, 8, 524. [Google Scholar] [CrossRef]
  40. Gómez-Pascual, A.; Rocamora-Pérez, G.; Ibanez, L.; Botía, J.A. Targeted co-expression networks for the study of traits. Sci. Rep. 2024, 14, 16675. [Google Scholar] [CrossRef]
  41. Gao, J.L.; Wang, Q.L.; Liu, J.; Zheng, S.Q.; Liu, J.H.; Gao, Z.Y.; Zhu, C. Integrating weighted gene co-expression network analysis and machine learning to elucidate neural characteristics in a mouse model of depression. Front. Psychiatry 2025, 16, 1564095. [Google Scholar] [CrossRef] [PubMed]
  42. Li, A.N.; Zhang, Y.R.; Zhao, Z.D.; Wang, M.M.; Zan, L.S. Molecular characterization and transcriptional regulation analysis of the bovine PDHB gene. PLoS ONE 2016, 11, e0157445. [Google Scholar] [CrossRef] [PubMed]
  43. de Azevedo, S.V.; Hartfelder, K. The insulin signaling pathway in honey bee (Apis mellifera) caste development—Differential expression of insulin-like peptides and insulin receptors in queen and worker larvae. J. Insect Physiol. 2008, 54, 1064–1071. [Google Scholar] [CrossRef] [PubMed]
  44. Sutherland, T.D.; Campbell, P.M.; Weisman, S.; Trueman, H.E.; Sriskantha, A.; Wanjura, W.J.; Haritos, V.S. A highly divergent gene cluster in honey bees encodes a novel silk family. Genome Res. 2006, 16, 1414–1421. [Google Scholar] [CrossRef]
  45. Xu, R.M.; Ma, B.B.; Yang, Y.Y.; Dong, X.C.; Li, J.K.; Xu, X.; Fang, Y. Proteome-metabolome profiling of wax gland complex reveals functional changes in honeybee, Apis mellifera L. iScience 2024, 27, 109279. [Google Scholar] [CrossRef]
  46. Robinson, E.K.; Jagannatha, P.; Covarrubias, S.; Cattle, M.; Smaliy, V.; Safavi, R.; Shapleigh, B.; Abu-Shumays, R.; Jain, M.; Cloonan, S.M.; et al. Inflammation drives alternative first exon usage to regulate immune genes including a novel iron-regulated isoform of Aim2. Elife 2021, 10, e69431. [Google Scholar] [CrossRef]
  47. Hasselmann, M.; Gempe, T.; Schiøtt, M.; Nunes-Silva, C.G.; Otte, M.; Beye, M. Evidence for the evolutionary nascence of a novel sex determination pathway in honeybees. Nature 2008, 454, 519–522. [Google Scholar] [CrossRef]
  48. Beye, M.; Hasselmann, M.; Fondrk, M.K.; Page, R.E.; Omholt, S.W. The gene Csd is the primary signal for sexual development in the honeybee and encodes an SR-type protein. Cell 2003, 114, 419–429. [Google Scholar] [CrossRef]
  49. Graveley, B.R. Sorting out the complexity of SR protein functions. RNA 2000, 6, 1197–1211. [Google Scholar] [CrossRef]
  50. Kay, B.K.; Williamson, M.P.; Sudol, M. The importance of being proline: The interaction of proline-rich motifs in signaling proteins with their cognate domains. FASEB J. 2000, 14, 231–241. [Google Scholar] [CrossRef] [PubMed]
  51. Butler, B.; Pirrotta, V.; Irminger-Finger, I.; Nöthiger, R. The sex-determining gene tra of Drosophila: Molecular cloning and transformation studies. EMBO J. 1986, 5, 3607–3613. [Google Scholar] [CrossRef]
  52. Boggs, R.T.; Gregor, P.; Idriss, S.; Belote, J.M.; McKeown, M. Regulation of sexual differentiation in D. melanogaster via alternative splicing of RNA from the transformer gene. Cell 1987, 50, 739–747. [Google Scholar] [CrossRef]
Genes 16 01409 g001
Figure 1. Principal component analysis (PCA) of the gene expression matrix for queens, workers, and drones (n = 6).
Figure 1. Principal component analysis (PCA) of the gene expression matrix for queens, workers, and drones (n = 6).
Genes 16 01409 g001
Genes 16 01409 g002
Figure 2. Differential gene expression analysis. (A) Statistical data of differentially expressed genes (D: Drone; W: Worker; Q: Queen). (B) Volcano Plots of Gene Expression in the Three Castes of Honeybee During 2-Day Larval Development (D: Drone; W: Worker; Q: Queen).
Figure 2. Differential gene expression analysis. (A) Statistical data of differentially expressed genes (D: Drone; W: Worker; Q: Queen). (B) Volcano Plots of Gene Expression in the Three Castes of Honeybee During 2-Day Larval Development (D: Drone; W: Worker; Q: Queen).
Genes 16 01409 g002
Genes 16 01409 g003
Figure 3. GO Analysis of Differentially Expressed Genes Among Three Castes of Honeybee (D: Drone; W: Worker; Q: Queen).
Figure 3. GO Analysis of Differentially Expressed Genes Among Three Castes of Honeybee (D: Drone; W: Worker; Q: Queen).
Genes 16 01409 g003
Genes 16 01409 g004
Figure 4. Construction of WGCNA Modules. (A) The distribution of gene numbers across the nine color modules. (B,C) Analysis of the scale–free topology fit index and average connectivity under different soft–thresholding powers. The red line indicates the soft–thresholding power of 24, where the correlation coefficient reaches 0.8. (D) Heatmap illustrating the correlation between gene expression and the three castes. The thickness of the lines indicates the strength of the correlation. Brown lines represent highly significant p-values, while green lines represent significant p-values. The stars highlight the correlations within each color module, with larger stars indicating stronger absolute correlation values.
Figure 4. Construction of WGCNA Modules. (A) The distribution of gene numbers across the nine color modules. (B,C) Analysis of the scale–free topology fit index and average connectivity under different soft–thresholding powers. The red line indicates the soft–thresholding power of 24, where the correlation coefficient reaches 0.8. (D) Heatmap illustrating the correlation between gene expression and the three castes. The thickness of the lines indicates the strength of the correlation. Brown lines represent highly significant p-values, while green lines represent significant p-values. The stars highlight the correlations within each color module, with larger stars indicating stronger absolute correlation values.
Genes 16 01409 g004
Genes 16 01409 g005
Figure 5. Training and validation performance of the TCN model. (A) Training and validation loss curves showing rapid convergence without overfitting. (B) Training and validation accuracy curves, with the model achieving 100% classification accuracy on the independent test set.
Figure 5. Training and validation performance of the TCN model. (A) Training and validation loss curves showing rapid convergence without overfitting. (B) Training and validation accuracy curves, with the model achieving 100% classification accuracy on the independent test set.
Genes 16 01409 g005
Genes 16 01409 g006
Figure 6. qPCR validation of RNA-seq data for caste-specific genes. Representative genes from drone (LOC408839), queen (LOC724161), and worker (LOC409153) showing concordant expression patterns between RNA-seq (left, n = 30/group) and qPCR (right, n = 5/group). Box plots show median, IQR, and mean ± SE (red diamonds with error bars). Group 0: caste-specific; Group 1: non-caste-specific. Statistical tests: Kruskal–Wallis (RNA-seq) and t-test (qPCR). *** p < 0.001 (Non-drone samples comprise pooled queens and workers at equal mass per replicate to test sex-level effects).
Figure 6. qPCR validation of RNA-seq data for caste-specific genes. Representative genes from drone (LOC408839), queen (LOC724161), and worker (LOC409153) showing concordant expression patterns between RNA-seq (left, n = 30/group) and qPCR (right, n = 5/group). Box plots show median, IQR, and mean ± SE (red diamonds with error bars). Group 0: caste-specific; Group 1: non-caste-specific. Statistical tests: Kruskal–Wallis (RNA-seq) and t-test (qPCR). *** p < 0.001 (Non-drone samples comprise pooled queens and workers at equal mass per replicate to test sex-level effects).
Genes 16 01409 g006
Genes 16 01409 g007
Figure 7. Alternative Splicing (AS) Events in the Three Castes of Honeybees. (A) The number of differentially spliced genes identified among the three honeybee castes (D: Drone; W: Worker; Q: Queen). (B) Types and distribution of alternative splicing events observed in the three castes of honeybees (D: Drone; W: Worker; Q: Queen). Alternative Splicing Events types: A3 (Alternative 3′ splice site, A3SS); A5 (Alternative 5′ splice site, A5SS); AF (Alternative First Exon, AF); AL (Alternative Last Exon, AL); MX (Mutually Exclusive Exons, MXE); RI (Retained Intron, RI); SE (Skipped Exon, SE).
Figure 7. Alternative Splicing (AS) Events in the Three Castes of Honeybees. (A) The number of differentially spliced genes identified among the three honeybee castes (D: Drone; W: Worker; Q: Queen). (B) Types and distribution of alternative splicing events observed in the three castes of honeybees (D: Drone; W: Worker; Q: Queen). Alternative Splicing Events types: A3 (Alternative 3′ splice site, A3SS); A5 (Alternative 5′ splice site, A5SS); AF (Alternative First Exon, AF); AL (Alternative Last Exon, AL); MX (Mutually Exclusive Exons, MXE); RI (Retained Intron, RI); SE (Skipped Exon, SE).
Genes 16 01409 g007
Genes 16 01409 g008
Figure 8. GO Enrichment Analysis of Alternative Splicing in the Three Castes of Honeybees.
Figure 8. GO Enrichment Analysis of Alternative Splicing in the Three Castes of Honeybees.
Genes 16 01409 g008
Genes 16 01409 g009
Figure 9. Sashimi Plots of Gene Splicing Events. (A) Sashimi plot showing the splicing patterns of the Csd gene. (B) Sashimi plot showing the splicing events of the Fem gene. (C) Sashimi plot showing the splicing patterns of the Tk gene. (Red: Drone; Blue: Queen; Green: Worker; Below each Sashimi plot are isoforms of the corresponding splicing event.The number on each arc indicates the number of reads that support the corresponding splicing junction.).
Figure 9. Sashimi Plots of Gene Splicing Events. (A) Sashimi plot showing the splicing patterns of the Csd gene. (B) Sashimi plot showing the splicing events of the Fem gene. (C) Sashimi plot showing the splicing patterns of the Tk gene. (Red: Drone; Blue: Queen; Green: Worker; Below each Sashimi plot are isoforms of the corresponding splicing event.The number on each arc indicates the number of reads that support the corresponding splicing junction.).
Genes 16 01409 g009
Genes 16 01409 g010
Figure 10. qPCR Validation of Caste-Specific Alternative Splicing in Sex Determination Genes. Expression levels of Csd, Fem, and Tk transcripts across drone, queen, and worker honeybees (n = 6 per caste). Box plots show median, interquartile range, and mean ± SE (red diamonds with error bars). Statistical significance was determined while employing ANOVA with Bonferroni correction (*** p < 0.001).
Figure 10. qPCR Validation of Caste-Specific Alternative Splicing in Sex Determination Genes. Expression levels of Csd, Fem, and Tk transcripts across drone, queen, and worker honeybees (n = 6 per caste). Box plots show median, interquartile range, and mean ± SE (red diamonds with error bars). Statistical significance was determined while employing ANOVA with Bonferroni correction (*** p < 0.001).
Genes 16 01409 g010
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Ding, X.; Li, J.; Yue, D.; Su, R. Integrating WGCNA, TCN, and Alternative Splicing to Map Early Caste Programs in Day-2 Honeybee Larvae. Genes 2025, 16, 1409. https://doi.org/10.3390/genes16121409

AMA Style

Ding X, Li J, Yue D, Su R. Integrating WGCNA, TCN, and Alternative Splicing to Map Early Caste Programs in Day-2 Honeybee Larvae. Genes. 2025; 16(12):1409. https://doi.org/10.3390/genes16121409

Chicago/Turabian Style

Ding, Xiang, Jinyou Li, Dan Yue, and Runlang Su. 2025. "Integrating WGCNA, TCN, and Alternative Splicing to Map Early Caste Programs in Day-2 Honeybee Larvae" Genes 16, no. 12: 1409. https://doi.org/10.3390/genes16121409

APA Style

Ding, X., Li, J., Yue, D., & Su, R. (2025). Integrating WGCNA, TCN, and Alternative Splicing to Map Early Caste Programs in Day-2 Honeybee Larvae. Genes, 16(12), 1409. https://doi.org/10.3390/genes16121409

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop