Expression analysis of candidate genes for fatty acid composition in adipose tissue and identification of regulatory regions

Genome-Wide Association Study Revealed Putative SNPs and Candidate Genes Associated with Growth and Meat Traits in Japanese Quail

The search for SNPs and candidate genes that determine the manifestation of major selected traits is one crucial objective for genomic selection aimed at increasing poultry production efficiency. Here, we report a genome-wide association study (GWAS) for traits characterizing meat performance in the domestic quail. A total of 146 males from an F2 reference population resulting from crossing a fast (Japanese) and a slow (Texas White) growing breed were examined. Using the genotyping-by-sequencing technique, genomic data were obtained for 115,743 SNPs (92,618 SNPs after quality control) that were employed in this GWAS. The results identified significant SNPs associated with the following traits at 8 weeks of age: body weight (nine SNPs), daily body weight gain (eight SNPs), dressed weight (33 SNPs), and weights of breast (18 SNPs), thigh (eight SNPs), and drumstick (three SNPs). Also, 12 SNPs and five candidate genes (GNAL, DNAJC6, LEPR, SPAG9, and SLC27A4) shared associations with three or more traits. These findings are consistent with the understanding of the genetic complexity of body weight-related traits in quail. The identified SNPs and genes can be used in effective quail breeding as molecular genetic markers for growth and meat characteristics for the purpose of genetic improvement.

Identification of genomic regions, genetic variants and gene networks regulating candidate genes for lipid metabolism in pig muscle

The intramuscular fat content and fatty acid composition of porcine meat have a significant impact on its quality and nutritional value. This research aimed to investigate the expression of 45 genes involved in lipid metabolism in the longissimus dorsi muscle of three experimental pig backcrosses, with a 25% of Iberian background. To achieve this objective, we conducted an expression Genome-Wide Association Study (eGWAS) using gene expression levels in muscle measured by high-throughput real-time qPCR for 45 target genes and genotypes from the PorcineSNP60 BeadChip or Axiom Porcine Genotyping Array and 65 single nucleotide polymorphisms (SNPs) located in 20 genes genotyped by a custom-designed Taqman OpenArray in a cohort of 354 animals. The eGWAS analysis identified 301 eSNPs associated with 18 candidate genes (ANK2, APOE, ARNT, CIITA, CPT1A, EGF, ELOVL6, ELOVL7, FADS3, FASN, GPAT3, NR1D2, NR1H2, PLIN1, PPAP2A, RORA, RXRA and UCP3). Three cis-eQTL (expression quantitative trait loci) were identified for GPAT3, RXRA, and UCP3 genes, which indicates that a genetic polymorphism proximal to the same gene is affecting its expression. Furthermore, 24 trans-eQTLs were detected, and eight candidate regulatory genes were located in these genomic regions. Additionally, two trans-regulatory hotspots in Sus scrofa chromosomes 13 and 15 were identified. Moreover, a co-expression analysis performed on 89 candidate genes and the fatty acid composition revealed the regulatory role of four genes (FABP5, PPARG, SCD, and SREBF1). These genes modulate the levels of α-linolenic, arachidonic, and oleic acids, as well as regulating the expression of other candidate genes associated with lipid metabolism. The findings of this study offer novel insights into the functional regulatory mechanism of genes involved in lipid metabolism, thereby enhancing our understanding of this complex biological process.

Genome-wide detection of CNV regions between Anqing six-end-white and Duroc pigs

BACKGROUND

Anqing six-end-white pig is a native breed in Anhui Province. The pigs have the disadvantages of a slow growth rate, low proportion of lean meat, and thick back fat, but feature the advantages of strong stress resistance and excellent meat quality. Duroc pig is an introduced pig breed with a fast growth rate and high proportion of lean meat. With the latter breed featuring superior growth characteristics but inferior meat quality traits, the underlying molecular mechanism that causes these phenotypic differences between Chinese and foreign pigs is still unclear.

RESULTS

In this study, copy number variation (CNV) detection was performed using the re-sequencing data of Anqing Six-end-white pigs and Duroc pigs, A total of 65,701 CNVs were obtained. After merging the CNVs with overlapping genomic positions, 881 CNV regions (CNVRs) were obtained. Based on the obtained CNVR information combined with their positions on the 18 chromosomes, a whole-genome map of the pig CNVs was drawn. GO analysis of the genes in the CNVRs showed that they were primarily involved in the cellular processes of proliferation, differentiation, and adhesion, and primarily involved in the biological processes of fat metabolism, reproductive traits, and immune processes.

CONCLUSION

The difference analysis of the CNVs between the Chinese and foreign pig breeds showed that the CNV of the Anqing six-end-white pig genome was higher than that of the introduced pig breed Duroc. Six genes related to fat metabolism, reproductive performance, and stress resistance were found in genome-wide CNVRs (DPF3, LEPR, MAP2K6, PPARA, TRAF6, NLRP4).

Deep Learning Technology Applied to Medical Image Tissue Classification

Medical image classification is a novel technology that presents a new challenge. It is essential that pathological images are automatically and correctly classified to enable doctors to provide precise treatment. Convolutional neural networks have demonstrated their effectiveness in classifying images in deep learning, which may have dozens or hundreds of layers, to illustrate the relationship between them in terms of their different neural network features. Convolutional layers consisting of small kernels take weights as input and guide them through an activation function as output. The main advantage of using convolutional neural networks (CNNs) instead of traditional neural networks is that they reduce the model parameters for greater accuracy. However, many studies have simply been focused on finding the best CNN model and classification results from a single medical image classification. Therefore, we applied a common deep learning network model in an attempt to identify the best model framework by training and validating different model parameters to classify medical images. After conducting experiments on six publicly available databases of pathological images, including colorectal cancer tissue, chest X-rays, common skin lesions, diabetic retinopathy, pediatric chest X-ray, and breast ultrasound image datasets, we were able to confirm that the recognition accuracy of the Inception V3 method was significantly better than that of other existing deep learning models.

Thermoelectric generator in endoreversible approximation: The effect of heat-transfer law under finite physical dimensions constraint

We revisit the optimal performance of a thermoelectric generator within the endoreversible approximation, while imposing a finite physical dimensions constraint in the form of a fixed total area of the heat exchangers. Our analysis is based on the linear-irreversible law for heat transfer between the reservoir and the working medium, in contrast to Newton's law usually assumed in literature. The optimization of power output is performed with respect to the thermoelectric current as well as the fractional area of the heat exchangers. We describe two alternate designs for allocating optimal areas to the heat exchangers. Interestingly, for each design, the use of linear-irreversible law yields the efficiency at maximum power in the well-known form 2η_{C}^{}/(4-η_{C}^{}), earlier obtained for the case of thermoelectric generator under exoreversible approximation, i.e., assuming only the internal irreversibility due to Joule heating. On the other hand, the use of Newton's law yields Curzon-Ahlborn efficiency.

Genome-wide identification of m6A-associated functional SNPs as potential functional variants for thyroid cancer

m6A methylation has been demonstrated to be one of the most important epigenetic regulation mechanisms in cell differentiation and cancer development especially m6A derived diagnostic and prognostic biomarkers have been identified in the past several years. However, systemic investigation to the interaction between germline single-nucleotide polymorphisms (SNPs) and m6A has not been conducted yet. In this study, we collected previous identified significant thyroid cancer associated SNPs from UKB cohort (358 cases and 407,399 controls) and ICR cohort (3,001 patients and 287,550 controls) and thyroid eQTL (sample size = 574 from GTEx project) and m6A-SNP (N = 1,678,126) were applied to prioritize the candidate SNPs. Finally, five candidate genes (PLEKHA8, SMUG1, CDC123, RMI2, ACSM5) were identified to be thyroid cancer associated m6A-related genetic susceptibility. Loss and gain function studies of m6A writer proteins confirm that ACSM5 is regulated by m6A methylation of mRNA. Moreover, ACSM5 is downregulated in thyroid cancer and inversely correlated with PTC malignancy and patient survival. Together, our study highlight mRNA-seq and m6A-seq double analysis provided a novel approach to identify cancer biomarkers and understanding the heterogeneity of human cancers.

Antipredator behavior in the rock-paper-scissors model

When faced with an imminent risk of predation, many animals react to escape consumption. Antipredator strategies are performed by individuals acting as a group to intimidate predators and minimize the damage when attacked. We study the antipredator prey response in spatial tritrophic systems with cyclic species dominance using the rock-paper-scissors game. The impact of the antipredator behavior is local, with the predation probability reducing exponentially with the number of prey in the predator's neighborhood. In contrast to the standard Lotka-Volterra implementation of the rock-paper-scissors model, where no spiral waves appear, our outcomes show that the antipredator behavior leads to spiral patterns from random initial conditions. The results show that the predation risk decreases exponentially with the level of antipredator strength. Finally, we investigate the coexistence probability and verify that antipredator behavior may jeopardize biodiversity for high mobility. Our findings may help biologists to understand ecosystems formed by species whose individuals behave strategically to resist predation.

Proteomic and lipidomic analyses reveal saturated fatty acids, phosphatidylinositol, phosphatidylserine, and associated proteins contributing to intramuscular fat deposition

Intramuscular fat (IMF) content is an important factor in porcine meat quality. Previous studies have screened multiple candidate genes related to IMF deposition, but the lipids that affect IMF deposition and their lipid-protein network remain unknown. In this study, we performed proteomic and lipidomic analyses of the longissimus dorsi (LD) muscle from high-IMF (IMFH) and low-IMF (IMF-L) groups of Xidu black pigs. Eighty-eight proteins and 143 lipids were differentially abundant between the groups. The differentially abundant proteins were found to be involved in cholesterol metabolism, the PPAR signaling pathway, and ferroptosis. The triacylglycerols (TAGs) upregulated in the IMF-H group were mainly shown to be synthesized by saturated fatty acids (SFAs), while the downregulated TAGs were mainly synthesized by polyunsaturated fatty acids (PUFAs). All differentially abundant phosphatidylinositols (PIs) and phosphatidylserines (PSs) were found to be upregulated in the IMF-H group. A correlation analysis of the proteomic and lipidomic revealed candidate proteins (APOA4, VDAC3, PRNP, CTSB, GSPT1) related to TAG, PI, and PS lipids. These results revealed differences in proteins and lipids between the IMF-H and IMF-L groups, which represent new candidate proteins and lipids that should be investigated to determine the molecular mechanisms controlling IMF deposition in pigs. SIGNIFICANCE: Intramuscular fat (IMF) is a key factor affecting meat quality, and meat with a higher IMF content can have a better flavor. In this study, proteomic results show that the ferroptosis pathway, including the PRNP, VDAC3 and CP proteins, affects IMF deposition. Lipid composition is the key factor affecting IMF deposition, but there are few reports on this. In this study, through lipidomic analysis, we suggest that saturated fatty acid (SFA), phosphatidylinositol (PI), and phosphatidylserine (PS) may contribute to IMF deposition. A correlation analysis reveals the potential regulatory network between lipids and proteins. This study clarifies the difference in protein and lipid compositions in longissimus dorsi (LD) muscle with high and low IMF contents. This information suggests that it would be beneficial to increase the intramuscular fat content of pork not only from a genetic perspective but also from a nutritional perspective.

Expression analysis of porcine miR-33a/b in liver, adipose tissue and muscle and its potential role in fatty acid metabolism

mir-33a and mir-33b are co-transcribed with the SREBF2 and SREBF1 transcription factors, respectively. The main role of SREBF1 is the regulation of genes involved in fatty acid metabolism, while SREBF2 regulates genes participating in cholesterol biosynthesis and uptake. Our objective was to study the expression of both miR-33a and miR-33b, together with their host SREBF genes, in liver, adipose tissue and muscle to better understand the role of miR-33a/b in the lipid metabolism of pigs. In our study, the expression of miR-33a, miR-33b and SREBF2 in liver, adipose tissue, and muscle was studied in 42 BC1_LD (25% Iberian x 75% Landrace backcross) pigs by RT-qPCR. In addition, the expression of in-silico predicted target genes and fatty acid composition traits were correlated with the miR-33a/b expression. We observed different tissue expression patterns for both miRNAs. In adipose tissue and muscle a high correlation between miR-33a and miR-33b expression was found, whereas a lower correlation was observed in liver. The expression analysis of in-silico predicted target-lipid related genes showed negative correlations between miR-33b and CPT1A expression in liver. Conversely, positive correlations between miR-33a and PPARGC1A and USF1 gene expression in liver were observed. Lastly, positive and negative correlations between miR-33a/b expression and saturated fatty acid (SFA) and polyunsaturated fatty acid (PUFA) content, respectively, were identified. Overall, our results suggested that both miRNAs are differentially regulated and have distinct functions in liver, in contrast to muscle and adipose tissue. Furthermore, the correlations between miR-33a/b expression both with the expression of in-silico predicted target-lipid related genes and with fatty acid composition, opens new avenues to explore the role of miR33a/b in the regulation of lipid metabolism.

Explanation and prediction of clinical data with imbalanced class distribution based on pattern discovery and disentanglement

Background

Statistical data analysis, especially the advanced machine learning (ML) methods, have attracted considerable interest in clinical practices. We are looking for interpretability of the diagnostic/prognostic results that will bring confidence to doctors, patients and their relatives in therapeutics and clinical practice. When datasets are imbalanced in diagnostic categories, we notice that the ordinary ML methods might produce results overwhelmed by the majority classes diminishing prediction accuracy. Hence, it needs methods that could produce explicit transparent and interpretable results in decision-making, without sacrificing accuracy, even for data with imbalanced groups.

Methods

In order to interpret the clinical patterns and conduct diagnostic prediction of patients with high accuracy, we develop a novel method, Pattern Discovery and Disentanglement for Clinical Data Analysis (cPDD), which is able to discover patterns (correlated traits/indicants) and use them to classify clinical data even if the class distribution is imbalanced. In the most general setting, a relational dataset is a large table such that each column represents an attribute (trait/indicant), and each row contains a set of attribute values (AVs) of an entity (patient). Compared to the existing pattern discovery approaches, cPDD can discover a small succinct set of statistically significant high-order patterns from clinical data for interpreting and predicting the disease class of the patients even with groups small and rare.

Results

Experiments on synthetic and thoracic clinical dataset showed that cPDD can 1) discover a smaller set of succinct significant patterns compared to other existing pattern discovery methods; 2) allow the users to interpret succinct sets of patterns coming from uncorrelated sources, even the groups are rare/small; and 3) obtain better performance in prediction compared to other interpretable classification approaches.

Conclusions

In conclusion, cPDD discovers fewer patterns with greater comprehensive coverage to improve the interpretability of patterns discovered. Experimental results on synthetic data validated that cPDD discovers all patterns implanted in the data, displays them precisely and succinctly with statistical support for interpretation and prediction, a capability which the traditional ML methods lack. The success of cPDD as a novel interpretable method in solving the imbalanced class problem shows its great potential to clinical data analysis for years to come.

Identification of strong candidate genes for backfat and intramuscular fatty acid composition in three crosses based on the Iberian pig

Meat quality has an important genetic component and can be modified by the fatty acid (FA) composition and the amount of fat contained in adipose tissue and muscle. The present study aimed to find genomic regions associated with the FA composition in backfat and muscle (longissimus dorsi) in 439 pigs with three different genetic backgrounds but having the Iberian breed in common. Genome-wide association studies (GWAS) were performed between 38,424 single-nucleotide polymorphisms (SNPs) covering the pig genome and 60 phenotypic traits related to backfat and muscle FA composition. Nine significant associated regions were found in backfat on the Sus scrofa chromosomes (SSC): SSC1, SSC2, SSC4, SSC6, SSC8, SSC10, SSC12, and SSC16. For the intramuscular fat, six significant associated regions were identified on SSC4, SSC13, SSC14, and SSC17. A total of 52 candidate genes were proposed to explain the variation in backfat and muscle FA composition traits. GWAS were also reanalysed including SNPs on five candidate genes (ELOVL6, ELOVL7, FADS2, FASN, and SCD). Regions and molecular markers described in our study may be useful for meat quality selection of commercial pig breeds, although several polymorphisms were breed-specific, and further analysis would be needed to evaluate possible causal mutations.

Identification of eQTLs associated with lipid metabolism in Longissimus dorsi muscle of pigs with different genetic backgrounds

Intramuscular fat content and its fatty acid composition affect porcine meat quality and its nutritional value. The present work aimed to identify genomic variants regulating the expression in the porcine muscle (Longissimus dorsi) of 45 candidate genes for lipid metabolism and fatty acid composition in three experimental backcrosses based on the Iberian breed. Expression genome-wide association studies (eGWAS) were performed between the muscle gene expression values, measured by real-time quantitative PCR, and the genotypes of 38,426 SNPs distributed along all chromosomes. The eGWAS identified 186 eSNPs located in ten Sus scrofa regions and associated with the expression of ACSM5, ACSS2, ATF3, DGAT2, FOS and IGF2 (FDR < 0.05) genes. Two expression quantitative trait loci (eQTLs) for IGF2 and ACSM5 were classified as cis-acting eQTLs, suggesting a mutation in the same gene affecting its expression. Conversely, ten eQTLs showed trans-regulatory effects on gene expression. When the eGWAS was performed for each backcross independently, only three common trans-eQTL regions were observed, indicating different regulatory mechanisms or allelic frequencies among the breeds. In addition, hotspot regions regulating the expression of several genes were detected. Our results provide new data to better understand the functional regulatory mechanisms of lipid metabolism genes in muscle.

Genome-wide detection of copy number variants in European autochthonous and commercial pig breeds by whole-genome sequencing of DNA pools identified breed-characterising copy number states

In this study, we identified copy number variants (CNVs) in 19 European autochthonous pig breeds and in two commercial breeds (Italian Large White and Italian Duroc) that represent important genetic resources for this species. The genome of 725 pigs was sequenced using a breed-specific DNA pooling approach (30-35 animals per pool) obtaining an average depth per pool of 42×. This approach maximised CNV discovery as well as the related copy number states characterising, on average, the analysed breeds. By mining more than 17.5 billion reads, we identified a total of 9592 CNVs (~683 CNVs per breed) and 3710 CNV regions (CNVRs; 1.15% of the reference pig genome), with an average of 77 CNVRs per breed that were considered as private. A few CNVRs were analysed in more detail, together with other information derived from sequencing data. For example, the CNVR encompassing the KIT gene was associated with coat colour phenotypes in the analysed breeds, confirming the role of the multiple copies in determining breed-specific coat colours. The CNVR covering the MSRB3 gene was associated with ear size in most breeds. The CNVRs affecting the ELOVL6 and ZNF622 genes were private features observed in the Lithuanian Indigenous Wattle and in the Turopolje pig breeds respectively. Overall, the genome variability unravelled here can explain part of the genetic diversity among breeds and might contribute to explain their origin, history and adaptation to a variety of production systems.

Relationships between osteoarthritic changes (osteophytes, porosity, eburnation) based on historical skeletal material

Background: Three main diagnostic types of osteoarthritic changes are distinguished in clinical and anthropological literature: osteophytes, porosity, and eburnation. The nature of the relationship between these changes and how lesions progress over time is still unclear.Aim: The aim of the present study is the analysis of the relationships between osteophytes, porosity, and eburnation based on skeletal material.Subjects and methods: The analysis employed the skeletal collection from Cedynia (199 individuals) from tenth to fourteenth-century Poland. Marginal osteophytes (OP), porosity (POR), and eburnation (EB) were examined on a shoulder, elbow, wrist, hip, knee, and ankle.Results: Osteophytes and porosity occurred independently of each other. Combinations of osteophytes and porosity (OP + POR) and osteophytes, porosity, and eburnation (OP + POR + EB) were rarely observed. Combinations of osteophytes and eburnation (OP + EB) or porosity and eburnation (POR + EB) were not found. There was a significant correlation between osteophytes and porosity in the scapula, proximal end of the ulna and proximal end of the femur. Osteophytes and eburnation were correlated at the distal end of the ulna. Porosity and eburnation were correlated at the distal end of the radius and distal end of the ulna. When all joints were considered together, all the types of osteoarthritic changes were correlated. However, the relationship between osteophytes and eburnation and between porosity and eburnation was only slightly significant. Osteophytes preceded porosity, but there were a few cases where more developed porosity accompanied less developed osteophytes.Conclusions: The findings indicate that correlations between osteoarthritic changes are weak, albeit statistically significant and further studies of the relationship between changes are necessary.

Insights into the metabolic responses of two contrasting Tibetan hulless barley genotypes under low nitrogen stress

Nitrogen (N) is an essential macronutrient for plants. However, excessive use of N fertilizer for cultivation is an environmental hazard. A good adaption to N deficiency is known in the Tibetan hulless barley. Therefore, it is of interest to complete the metabolic analysis on LSZQK which is a low nitrogen (low-N) sensitive genotype and Z0284 that is tolerant to low-N. We identified and quantified 750 diverse metabolites in this analysis. The two genotypes show differences in their basal metabolome under normal N condition. Polyphenols and lipids related metabolites were significantly enriched in Z0284 having a basal role prior to exposure to low-N stress. Analysis of the differentially accumulated metabolites (DAM) induced by low-N explain the genotype-specific responses. Fourteen DAMs showed similar patterns of change between low-N and control conditions in both genotypes. This could be the core low-N responsive metabolites regardless of the tolerance level in hulless barley. We also identified 4 DAMs (serotonin, MAG (18:4) isomer 2, tricin 7-O-feruloylhexoside and gluconic acid) shared by both genotypes displaying opposite patterns of regulation under low-N conditions and may play important roles in low-N tolerance. This report provides a theoretical basis for further understanding of the molecular mechanisms of low-N stress tolerance in hulless barley.

Genome-wide association study for carcass quality traits and growth in purebred and crossbred pigs1

Carcass quality traits such as back fat (BF), loin depth (LD), and ADG are of extreme economic importance for the swine industry. This study aimed to (i) estimate the genetic parameters for such traits and (ii) conduct a single-step genome-wide association study (ssGWAS) to identify genomic regions that affect carcass quality and growth traits in purebred (PB) and three-way crossbred (CB) pigs. A total of 28,497 PBs and 135,768 CBs pigs were phenotyped for BF, LD, and ADG. Of these, 4,857 and 3,532 were genotyped using the Illumina PorcineSNP60K Beadchip. After quality control, 36,328 SNPs were available and were used to perform an ssGWAS. A bootstrap analysis (n = 1,000) and a signal enrichment analysis were performed to declare SNP significance. Genome regions were based on the variance explained by significant 10-SNP sliding windows. Estimates of PB heritability (SE) were 0.42 (0.019) for BF, 0.39 (0.020) for LD, and 0.35 (0.021) for ADG. Estimates of CB heritability were 0.49 (0.042) for BF, 0.27 (0.029) for LD, and 0.12 (0.021) for ADG. Genetic correlations (SE) across the two populations were 0.81 (0.02), 0.79 (0.04), and 0.56 (0.05), for BF, LD, and ADG, respectively. The variance explained by significant regions for each trait in PBs ranged from 1.51% to 1.35% for BF, from 4.02% to 3.18% for LD, and from 2.26% to 1.45% for ADG. In CBs, the variance explained by significant regions ranged from 1.88% to 1.37% for BF, from 1.29% to 1.23% for LD, and from 1.54% to 1.32% for ADG. In this study, we have described regions of the genome that determine carcass quality and growth traits of PB and CB pigs. These results provide evidence that there are overlapping and nonoverlapping regions in the genome influencing carcass quality and growth traits in PBs and three-way CB pigs.

Plasma Metabolites Associated with Frequent Red Wine Consumption: A Metabolomics Approach within the PREDIMED Study

SCOPE

The relationship between red wine (RW) consumption and metabolism is poorly understood. It is aimed to assess the systemic metabolomic profiles in relation to frequent RW consumption as well as the ability of a set of metabolites to discriminate RW consumers.

METHODS AND RESULTS

A cross-sectional analysis of 1157 participants is carried out. Subjects are divided as non-RW consumers versus RW consumers (>1 glass per day RW [100 mL per day]). Plasma metabolomics analysis is performed using LC-MS. Associations between 386 identified metabolites and RW consumption are assessed using elastic net regression analysis taking into consideration baseline significant covariates. Ten-cross-validation (CV) is performed and receiver operating characteristic curves are constructed in each of the validation datasets based on weighted models. A subset of 13 metabolites is consistently selected and RW consumers versus nonconsumers are discriminated. Based on the multi-metabolite model weighted with the regression coefficients of metabolites, the area under the curve is 0.83 (95% CI: 0.80-0.86). These metabolites mainly consisted of lipid species, some organic acids, and alkaloids.

CONCLUSIONS

A multi-metabolite model identified in a Mediterranean population appears useful to discriminate between frequent RW consumers and nonconsumers. Further studies are needed to assess the contribution of these metabolites in health and disease.

Machine learning applied to transcriptomic data to identify genes associated with feed efficiency in pigs

Background

To date, the molecular mechanisms that underlie residual feed intake (RFI) in pigs are unknown. Results from different genome-wide association studies and gene expression analyses are not always consistent. The aim of this research was to use machine learning to identify genes associated with feed efficiency (FE) using transcriptomic (RNA-Seq) data from pigs that are phenotypically extreme for RFI.

Methods

RFI was computed by considering within-sex regression on mean metabolic body weight, average daily gain, and average backfat gain. RNA-Seq analyses were performed on liver and duodenum tissue from 32 high and 33 low RFI pigs collected at 153 d of age. Machine-learning algorithms were used to predict RFI class based on gene expression levels in liver and duodenum after adjusting for batch effects. Genes were ranked according to their contribution to the classification using the permutation accuracy importance score in an unbiased random forest (RF) algorithm based on conditional inference. Support vector machine, RF, elastic net (ENET) and nearest shrunken centroid algorithms were tested using different subsets of the top rank genes. Nested resampling for hyperparameter tuning was implemented with tenfold cross-validation in the outer and inner loops.

Results

The best classification was obtained with ENET using the expression of 200 genes in liver [area under the receiver operating characteristic curve (AUROC): 0.85; accuracy: 0.78] and 100 genes in duodenum (AUROC: 0.76; accuracy: 0.69). Canonical pathways and candidate genes that were previously reported as associated with FE in several species were identified. The most remarkable pathways and genes identified were NRF2-mediated oxidative stress response and aldosterone signalling in epithelial cells, the DNAJC6, DNAJC1, MAPK8, PRKD3 genes in duodenum, and melatonin degradation II, PPARα/RXRα activation, and GPCR-mediated nutrient sensing in enteroendocrine cells and SMOX, IL4I1, PRKAR2B, CLOCK and CCK genes in liver.

Conclusions

ML algorithms and RNA-Seq expression data were found to provide good performance for classifying pigs into high or low RFI groups. Classification was better with gene expression data from liver than from duodenum. Genes associated with FE in liver and duodenum tissue that can be used as predictive biomarkers for this trait were identified.

Electronic supplementary material

The online version of this article (10.1186/s12711-019-0453-y) contains supplementary material, which is available to authorized users.

Effect of dietary polyunsaturated fatty acid and antioxidant supplementation on the transcriptional level of genes involved in lipid and energy metabolism in swine

Porcine fat traits depend mostly on the interaction between nutritional and genetic factors. However, the pathways and biological processes influenced by this interaction are still poorly known in pigs, although they can have a huge impact on meat quality traits. The present research provides new knowledge insight into the effect of four diets (D1 = standard diet; D2 = linseed supplementation; D3 = linseed, vitamin E and selenium supplementation; D4 = linseed and plant-derived polyphenols supplementation) on the expression of 24 candidate genes selected for their role in lipid and energy metabolism. The data indicated that 10 out of 24 genes were differentially expressed among diets, namely ACACA, ADIPOQ, ADIPOR1, CHREBP (MLXPL), ELOVL6, FASN, G6PD, PLIN2, RXRA and SCD. Results from the univariate analysis displayed an increased expression of ACACA, ADIPOQ, ADIPOR1, CHREBP, ELOVL6, FASN, PLIN2, RXRA and SCD in D4 compared to D2. Similarly, ACACA, ADIPOQ, ADIPOR1, ELOVL6 and SCD were highly expressed in D4 compared to D3, while no differences were observed in D2-D3 comparison. Moreover, an increased expression of G6PD and ELOVL6 genes in D4 compared to D1 was observed. Results from the multivariate analysis confirmed that D2 was not different from D3 and that ACACA, SCD and FASN expression made D4 different from D2 and D3. Comparing D4 and D1, the expression levels of ELOVL6 and ACACA were the most influenced. This research provides evidence that the addition of both n-3 PUFA and polyphenols, derived from linseed, grape-skin and oregano supplementation in the diets, stimulates the expression of genes involved in lipogenesis and in oxidative processes. Results evidenced a greater effect on gene expression of the diet added with both plant extracts and n-3 PUFA, resulting in an increased expression of genes coding for fatty acid synthesis, desaturation and elongation in pig Longissimus thoracis muscle.

A polymorphism in the fatty acid desaturase-2 gene is associated with the arachidonic acid metabolism in pigs

Arachidonic acid (C20:4) is related to a wide range of biological effects including lipid homeostasis. The fatty acid desaturase-2 (FADS2) gene encodes for the delta-6-desaturase, which is involved in the biosynthesis of C20:4 from linoleic acid (C18:2). The purpose of this study was to characterise mutations in the promoter of the porcine FADS2, evaluating in particular the effect of one haplotype tagging polymorphism (rs321384923A > G) on the biosynthesis pathway of C20:4. A total of 1,192 Duroc barrows with records on fatty acid composition in muscle and subcutaneous fat were genotyped. Pigs carrying the A allele showed, irrespective of fat content, both enhanced FADS2 expression and higher C20:4 in muscle and exhibited increased ratios of C20:4 to C18:2 and of C20:4 to eicosadienoic acid (C20:2) in both muscle and adipose tissue. Despite the inverse relationship observed between C20:4 and fat content, the rs321384923 polymorphism had no impact on lean weight. It is concluded that the haplotype encompassing the rs321384923 polymorphism at the porcine FADS2 affects the n-6 fatty acid profile by specifically modifying the desaturation efficiency of C18:2 to C20:4 rather than by concomitant variations in C18:2 following changes in fat content.