Residual network improves the prediction accuracy of genomic selection.

Genetic improvement of complex traits in animal and plant breeding depends on the efficient and accurate estimation of breeding values. Deep learning methods have been shown to be not superior over traditional genomic selection (GS) methods, partially due to the degradation problem (i.e. with the increase of the model depth, the performance of the deeper model deteriorates). Since the deep learning method residual network (ResNet) is designed to solve gradient degradation, we examined its performance and factors related to its prediction accuracy in GS. Here we compared the prediction accuracy of conventional genomic best linear unbiased prediction, Bayesian methods (BayesA, BayesB, BayesC, and Bayesian Lasso), and two deep learning methods, convolutional neural network and ResNet, on three datasets (wheat, simulated and real pig data). ResNet outperformed other methods in both Pearson's correlation coefficient (PCC) and mean squared error (MSE) on the wheat and simulated data. For the pig backfat depth trait, ResNet still had the lowest MSE, whereas Bayesian Lasso had the highest PCC. We further clustered the pig data into four groups and, on one separated group, ResNet had the highest prediction accuracy (both PCC and MSE). Transfer learning was adopted and capable of enhancing the performance of both convolutional neural network and ResNet. Taken together, our findings indicate that ResNet could improve GS prediction accuracy, affected potentially by factors such as the genetic architecture of complex traits, data volume, and heterogeneity.

Valine induces inflammation and enhanced adipogenesis in lean mice by multi-omics analysis.

Introduction: The branched-chain amino acids (BCAAs) are essential to mammalian growth and development but aberrantly elevated in obesity and diabetes. Each BCAA has an independent and specific physio-biochemical effect on the host. However, the exact molecular mechanism of the detrimental effect of valine on metabolic health remains largely unknown. Methods and results: This study showed that for lean mice treated with valine, the hepatic lipid metabolism and adipogenesis were enhanced, and the villus height and crypt depth of the ileum were significantly increased. Transcriptome profiling on white and brown adipose tissues revealed that valine disturbed multiple signaling pathways (e.g., inflammation and fatty acid metabolism). Integrative cecal metagenome and metabolome analyses found that abundances of Bacteroidetes decreased, but Proteobacteria and Helicobacter increased, respectively; and 87 differential metabolites were enriched in several molecular pathways (e.g., inflammation and lipid and bile acid metabolism). Furthermore, abundances of two metabolites (stercobilin and 3-IAA), proteins (AMPK/pAMPK and SCD1), and inflammation and adipogenesis-related genes were validated. Discussion: Valine treatment affects the intestinal microbiota and metabolite compositions, induces gut inflammation, and aggravates hepatic lipid deposition and adipogenesis. Our findings provide novel insights into and resources for further exploring the molecular mechanism and biological function of valine on lipid metabolism.

Triangular pulley traction facilitates endoscopic full-thickness resection of exogenous gastric tumor in the fornix of the stomach.

A 65-year-old woman was diagnosed with an exogenous submucosal tumor located in the fornix of the stomach, on the basis of the endoscopic ultrasound and enhanced CT findings. She refused surgery and referred for EFTR. It is difficult to perform EFTR at the gastric fornix and suture the large surgical defect. Therefore, we created technique of triangular pulley traction combined with pre-closure.

l-valine supplementation disturbs vital molecular pathways and induces apoptosis in mouse testes.

The branched-chain amino acids (BCAAs: leucine, isoleucine and valine) are essential for animal growth and metabolic health. However, the effect of valine on male reproduction and its underlying molecular mechanism remain largely unknown. Here, we showed that l-valine supplementation (0.30% or 0.45%, water drinking for 3 weeks) did not change body and testis weights, but significantly altered morphology of sertoli cells and germ cells within seminiferous tubule, and enlarged the space between seminiferous tubules within mouse testis. l-valine treatment (0.45%) increased significantly the Caspase3/9 mRNA levels and CASPASE9 protein levels, therefore induced apoptosis of mouse testis. Moreover, gene expression levels related to autophagy (Atg5 and Lamb3), DNA 5 mC methylation (Dnmt1, Dnmt3a, Tet2 and Tet3), RNA m6A methylation (Mettl14, Alkbh5 and Fto), and m6A methylation binding proteins (Ythdf1/2/3 and Igf2bp1/2) were significantly reduced. Protein abundances of ALKBH5, FTO and YTHDF3 were also significantly reduced, but not for ATG5 and TET2. Testis transcriptome sequencing detected 537 differentially expressed genes (DEGs, 26 up-regulated and 511 down-regulated), involved in multiple important signaling pathways. RT-qPCR validated 8 of 9 DEGs (Cd36, Scd1, Insl3, Anxa5, Lcn2, Hsd17b3, Cyp11a1, Cyp17a1 and Agt) to be decreased significantly, consistent with RNA-seq results. Taken together, l-valine treatment could disturb multiple signaling pathways (autophagy and RNA methylation etc.), and induce apoptosis to destroy the tissue structure of mouse testis.

Enantioselective α-Amination of Amides by One-Pot Organo-/Iodine Sequential Catalysis.

An one-pot organo- and iodine sequential catalysis strategy for reactions of amides with pyrazole-based primary amines was described to synthesize chiral α-amino amides with a quaternary stereocenter. This methodology exhibited strong asymmetric induction, resulting in a typical enantiomeric excess value exceeding 99% and diastereoselectivity up to >99:1 dr. Moreover, the reaction was conducted without the use of any metals or strong bases.

Entire traction using clip-and-nylon ring to facilitate endoscopic submucosal dissection of a laterally spreading tumor with fibrosis in the rectum.

A 78-year-old woman with hematochezia underwent a colonoscopy and found a 2 × 2-cm laterally spreading tumor (LST) in the rectum, 3 cm from the anus. Because of the risk related to anus preservation and the potential operative trauma, the patient refused surgery and was referred for ESD treatment. Here, we applied a novel entire traction method to deal with this subset of tumors.

Integrated single cell transcriptome sequencing analysis reveals species-specific genes and molecular pathways for pig spermiogenesis.

Mammalian spermatogenesis is a highly complicated and intricately organized process involving spermatogonia propagation (mitosis) and meiotic differentiation into mature sperm cells (spermiogenesis). In pigs, spermatogonia development and the role of somatic cells in spermatogenesis were previously investigated in detail. However, the characterization of key molecules fundamental to pig spermiogenesis remains less explored. Here we compared spermatogenesis between humans and pigs, focusing on spermiogenesis, by integrative testicular single-cell RNA sequencing (scRNA-seq) analysis. Human and pig testicular cells were clustered into 26 different groups, with cell-type-specific markers and signalling pathways. For spermiogenesis, pseudo-time analysis classified the lineage differentiation routes for round, elongated spermatids and spermatozoa. Moreover, markers and molecular pathways specific to each type of spermatids were examined for humans and pigs, respectively. Furthermore, high-dimensional weighted gene co-expression network analysis (hdWGCNA) identified gene modules specific for each type of human and pig spermatids. Hub genes (pig: SNRPD2.1 related to alternative splicing; human: CATSPERZ, Ca[2+] ion channel) potentially involved in spermiogenesis were also revealed. Taken together, our integrative analysis found that human and pig spermiogeneses involve specific genes and molecular pathways and provided resources and insights for further functional investigation on spermatid maturation and male reproductive ability.

Advances in the discovery of genetic elements underlying longissimus dorsi muscle growth and development in the pig.

As a major source of protein in human diets, pig meat plays a crucial role in ensuring global food security. Key determinants of meat production refer to the chemical and physical compositions or characteristics of muscle fibers, such as the number, hypertrophy potential, fiber-type conversion and intramuscular fat deposition. However, the growth and formation of muscle fibers comprises a complex process under spatio-temporal regulation, that is, the intermingled and concomitant proliferation, differentiation, migration and fusion of myoblasts. Recently, with the fast and continuous development of next-generation sequencing technology, the integration of quantitative trait loci mapping with genome-wide association studies (GWAS) has greatly helped animal geneticists to discover and explore thousands of functional or causal genetic elements underlying muscle growth and development. However, owing to the underlying complex molecular mechanisms, challenges to in-depth understanding and utilization remain, and the cost of large-scale sequencing, which requires integrated analyses of high-throughput omics data, is high. In this review, we mainly elaborate on research advances in integrative analyses (e.g. GWAS, omics) for identifying functional genes or genomic elements for longissimus dorsi muscle growth and development for different pig breeds, describing several successful transcriptome analyses and functional genomics cases, in an attempt to provide some perspective on the future functional annotation of genetic elements for muscle growth and development in pigs.

Single-cell RNA sequencing reveals blastomere heterogeneity of 2-cell embryos in pigs.

In mammals, single blastomeres from as early as 2-cell embryos demonstrate heterogeneous developmental capacity and fate decision into different cell lineages. However, mechanisms underlying blastomere heterogeneity of 2-cell embryos remain largely unresolved. Here, we analysed the molecular heterogeneity of full-length mRNAs and their 3'UTR regions, based on the single-cell RNA-seq data of pig 2-cell embryos generated from in vivo fertilization (in vivo), in vitro fertilization (in vitro) and parthenogenetic activation (PA), respectively. First, unsupervised clustering helped discover two different groups of blastomeres for 2-cell pig embryos. Between these two groups of blastomeres in pig 2-cell embryos, 35, 301 and 428 full-length mRNAs respectively in in vivo, in vitro and PA embryo types were identified to be differentially expressed (padj ≤ .05 and |log2 [fold change]| ≥1) (DE mRNAs), while 92, 89 and 42 mRNAs were shown to be with significantly different 3'UTR lengths (3'UTR DE) (padj ≤ .05). Gene enrichment for both DE mRNAs and 3'UTR DE mRNAs found multiple signalling pathways, including cell cycle, RNA processing. Few numbers of common DE mRNAs and 3'UTR DE mRNAs existed between in vitro and in vivo blastomeres derived from 2-cell embryos, indicating the larger differences between in vitro and in vivo fertilized embryos. Integrative genomics viewer analysis further identified that 3'UTRs of HSDL2 and SGTA (in vivo), FAM204A and phosphoserine phosphatase (in vitro), PRPF40A and RPIA (PA) had >100 nt average length changes. Moreover, numbers and locations of regulatory elements (polyadenylation site, cytoplasmic polyadenylation element and microRNA binding sites) within 3'UTRs of these DE mRNAs were predicted. These results indicate that molecular heterogeneity existed among blastomeres from different types of pig 2-cell embryos, providing useful information and novel insights into future functional investigation on its relationship with the subsequent embryo development and differentiation.

GWLD: an R package for genome-wide linkage disequilibrium analysis.

Linkage disequilibrium (LD) analysis is fundamental to the investigation of the genetic architecture of complex traits (e.g. human disease, animal and plant breeding) and population structure and evolution dynamics. However, until now, studies primarily focus on LD status between genetic variants located on the same chromosome. Moreover, genome (re)sequencing produces unprecedented numbers of genetic variants, and fast LD computation becomes a challenge. Here, we have developed GWLD, a parallelized and generalized tool designed for the rapid genome-wide calculation of LD values, including conventional D/D', r2, and (reduced) mutual information (MI and RMI) measures. LD between genetic variants within and across chromosomes can be rapidly computed and visualized in either an R package or a standalone C++ software package. To evaluate the accuracy and speed of LD calculation, we conducted comparisons using 4 real datasets. Interchromosomal LD patterns observed potentially reflect levels of selection intensity across different species. Both versions of GWLD, the R package (https://github.com/Rong-Zh/GWLD/GWLD-R) and the standalone C++ software (https://github.com/Rong-Zh/GWLD/GWLD-C++), are freely available on GitHub.

Dynamic changes of 3'UTR length during oocyte-to-zygote transition of in vitro pig embryos.

Alternative polyadenylation (APA) generates different 3'-untranslated regions (3'UTRs) to regulate gene expression and localization, and affects a variety of biological processes. Here, we characterized the 3'UTR dynamics during the oocyte-to-zygote transition by analysing our previously reported porcine single-cell RNA-seq (scRNA-seq) datasets (in vitro matured metaphase II (MII) oocytes, in vitro fertilized zygotes (IVF1) and parthenogenetically activated 1-cell embryos (PA1)). After IVF1 versus MII comparison, dynamic analyses of APA from RNA-seq (DaPars) method identified 139 mRNAs with significantly different 3'UTRs (padj . ≤ .05), mainly enriched in cell cycle, regulation of cyclin-dependent protein kinase activity, histone modification, mRNA surveillance, and regulation of actin cytoskeleton. For PA1 versus MII comparison, 105 mRNAs with significantly different 3'UTRs (padj . ≤ .05) were identified to be mainly enriched in intracellular transport, mitotic spindle organization, cell cycle, pyruvate metabolism and glycolysis/gluconeogenesis. Furthermore, there were 7 mRNAs with more significant 3'UTR differences (|△PDUI| ≥ 0.45 and |log2 [PDUI ratio]| ≥ 0.59) respectively in IVF1 versus MII (Lrp2bp, Mtfr2, Nhlrc2, Psip1, Smu1, Ssr1 and Wtap) and PA1 versus MII (Asf1b, Dimt1, Nap1l1, Ncoa4, Nudt21, Pnn and Rpl15) comparisons. Integrative genomics viewer analysis further identified that 3'UTRs of Psip1, Smu1, Ssr1 and Wtap had more than 140 nt average length changes, whereas those of Dimt1, Nap1l1 and Rpl15 were shortened with more than 460 nt. Regulatory elements (PAS, CPE, microRNA binding sites and m6 A sites) in 3'UTRs of different lengths were predicted. Our findings provide useful information to further investigate the molecular mechanism of 3'UTR in regulating the oocyte-to-zygote transition of pig embryos.

Selection of reference genes for RT-qPCR analysis in developing chicken embryonic ovary.

BACKGROUND: Normalization of the expression profiling of target genes, in a tissue-specific manner and under different experimental conditions, requires stably expressed gene(s) to be used as internal reference(s). However, to study the molecular regulation of oocyte meiosis initiation during ovary development in chicken embryos, stable reference gene(s) still need to be compared and confirmed. METHODS AND RESULTS: Six candidate genes previously used as internal references for the chicken embryo (Actb, Cvh, Dazl, Eef1a, Gapdh and Rpl15) were chosen, and their expression profiles in left ovaries dissected at five chicken embryonic days (E12.5, E15.5, E17.5, E18.5 and E20.5) were evaluated, respectively. Separately, GeNorm, NormFinder, BestKeeper and Comparative ΔCt methods were used to assess the stability of candidate reference genes, and all results were combined to give the final rank by RefFinder. All methods identified that Eef1a and Rpl15 were the two most stable internal reference genes, whereas Cvh is the most unstable one. Moreover, expression levels of three marker genes for chicken oocyte meiosis entry (Stra8, Scp3 and Dmc1) were normalized, based on Eef1a, Rpl15, or their combinations, respectively. CONCLUSION: Our findings provide the most suitable internal reference genes (Eef1a and Rpl15), to investigate further molecular regulation of ovary development and oocyte meiosis initiation in chicken embryos.