Surgically castrated male pigs, which are commonly produced in pork production systems, have slightly lower production efficiency, compared to intact female pigs (gilts). This is mainly due to an unfavorable feed conversion rate and fatter carcasses. These differences influenced by physiological and genetic factors can be identified through metabolomics, which describes metabolic profiles. In this study, we used untargeted metabolomics to analyze the plasma of 694 Italian Large White pigs (228 castrated males and 466 intact gilts), sampled at slaughter. The metabolomic profiles included 731 metabolites covering 98 sub-pathways. The raw metabolomic data were cleaned and imputed using multivariate imputation by chained equations. The Boruta algorithm was then employed to identify metabolites that have different concentrations between castrated males and intact gilts. To address the random nature of feature selection, multiple Boruta runs were generated, and nested within a 10-fold cross-validation, resulting in 1,250 Boruta datasets. These datasets helped identify 40 informative metabolites, with a reduced core of 15 metabolites consistently confirmed across all runs. Their calculated random forest out-of-bag error was 0.25 and 0.27, respectively. The relevance, ranking, and predictive ability of each selected metabolite were determined based on the mean decrease Gini (MDG) and the area under the curve (AUC) of the receiver operating characteristic curve analysis, with MDG values of 0.024 ± 0.007 and 0.030 ± 0.009 and AUC values of 0.62 ± 0.04 and 0.65 ± 0.03 for the 2 metabolite sets, respectively. Of the 40 selected metabolites, 60% had higher concentrations in castrated males than in intact gilts, while in the 15 metabolites set, this percentage was 80%. Network and biological pathways analyses indicated that the selected metabolites were primarily amino acids and lipids, many of which belonged to their respective sub-pathways, suggesting minimal biological differences between castrated males and intact gilts. These findings support previous results obtained using a targeted metabolomic platform. This study represents the largest investigation to date on the pig sex metabolome, providing essential biological insights that could inform precise husbandry and feeding strategies in pigs, taking into consideration the castration status of the males.
Bovo, S., Bolner, M., Schiavo, G., Galimberti, G., Bertolini, F., Dall'Olio, S., et al. (2025). Description of metabolic differences between castrated males and intact gilts obtained from high-throughput metabolomics of porcine plasma. JOURNAL OF ANIMAL SCIENCE, 103, 1-17 [10.1093/jas/skaf178].
Description of metabolic differences between castrated males and intact gilts obtained from high-throughput metabolomics of porcine plasma
Bovo, Samuele;Bolner, Matteo;Schiavo, Giuseppina;Galimberti, Giuliano;Bertolini, Francesca;Dall'Olio, Stefania;Ribani, Anisa;Zambonelli, Paolo;Fontanesi, Luca
2025
Abstract
Surgically castrated male pigs, which are commonly produced in pork production systems, have slightly lower production efficiency, compared to intact female pigs (gilts). This is mainly due to an unfavorable feed conversion rate and fatter carcasses. These differences influenced by physiological and genetic factors can be identified through metabolomics, which describes metabolic profiles. In this study, we used untargeted metabolomics to analyze the plasma of 694 Italian Large White pigs (228 castrated males and 466 intact gilts), sampled at slaughter. The metabolomic profiles included 731 metabolites covering 98 sub-pathways. The raw metabolomic data were cleaned and imputed using multivariate imputation by chained equations. The Boruta algorithm was then employed to identify metabolites that have different concentrations between castrated males and intact gilts. To address the random nature of feature selection, multiple Boruta runs were generated, and nested within a 10-fold cross-validation, resulting in 1,250 Boruta datasets. These datasets helped identify 40 informative metabolites, with a reduced core of 15 metabolites consistently confirmed across all runs. Their calculated random forest out-of-bag error was 0.25 and 0.27, respectively. The relevance, ranking, and predictive ability of each selected metabolite were determined based on the mean decrease Gini (MDG) and the area under the curve (AUC) of the receiver operating characteristic curve analysis, with MDG values of 0.024 ± 0.007 and 0.030 ± 0.009 and AUC values of 0.62 ± 0.04 and 0.65 ± 0.03 for the 2 metabolite sets, respectively. Of the 40 selected metabolites, 60% had higher concentrations in castrated males than in intact gilts, while in the 15 metabolites set, this percentage was 80%. Network and biological pathways analyses indicated that the selected metabolites were primarily amino acids and lipids, many of which belonged to their respective sub-pathways, suggesting minimal biological differences between castrated males and intact gilts. These findings support previous results obtained using a targeted metabolomic platform. This study represents the largest investigation to date on the pig sex metabolome, providing essential biological insights that could inform precise husbandry and feeding strategies in pigs, taking into consideration the castration status of the males.| File | Dimensione | Formato | |
|---|---|---|---|
|
skaf178.pdf
accesso aperto
Tipo:
Versione (PDF) editoriale / Version Of Record
Licenza:
Licenza per Accesso Aperto. Creative Commons Attribuzione (CCBY)
Dimensione
3.08 MB
Formato
Adobe PDF
|
3.08 MB | Adobe PDF | Visualizza/Apri |
|
skaf178_suppl_supplementary_tables_s1-s3_figure_s1.docx
accesso aperto
Descrizione: Supplementary Data
Tipo:
File Supplementare
Licenza:
Licenza per Accesso Aperto. Creative Commons Attribuzione (CCBY)
Dimensione
176.43 kB
Formato
Microsoft Word XML
|
176.43 kB | Microsoft Word XML | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


