Search

Single-cell RNA sequencing (scRNA-seq) is now a successful technology for identifying cell heterogeneity, revealing new cell subpopulations, and predicting developmental trajectories. A crucial component in scRNA-seq is the precise identification of cell subsets. Although many unsupervised clustering methods have been developed for clustering cell subpopulations, the performance of these methods is prone to be affected by dropout, high dimensionality, and technical noise. Additionally, most existing methods are time-consuming and fail to fully consider the potential correlations between cells. In this paper, we propose a novel unsupervised clustering method called scVGATAE (Single-cell Variational Graph Attention Autoencoder) for scRNA-seq data. This method constructs a reliable cell graph through network denoising, utilizes a novel variational graph autoencoder model integrated with graph attention networks to aggregate neighbor information and learn the distribution of the low-dimensional representations of cells, and adaptively determines the model training iterations for various datasets. Finally, the obtained low-dimensional representations of cells are clustered using kmeans. Experiments on nine public datasets show that scVGATAE outperforms classical and state-of-the-art clustering methods.

Body size may drive the molecular evolution of mitochondrial genes in response to changes in energy requirements across species of different sizes. In this study, we perform selection pressure analysis and phylogenetic independent contrasts (PIC) to investigate the association between molecular evolution of mitochondrial genome protein-coding genes (mtDNA PCGs) and body size in terrestrial Cetartiodactyla. Employing selection pressure analysis, we observe that the average non-synonymous/synonymous substitution rate ratio (ω) of mtDNA PCGs is significantly reduced in small-bodied species relative to their medium and large counterparts. PIC analysis further confirms that ω values are positively correlated with body size (R2 = 0.162, p = 0.0016). Our results suggest that mtDNA PCGs of small-bodied species experience much stronger purifying selection as they need to maintain a heightened metabolic rate. On the other hand, larger-bodied species may face less stringent selective pressures on their mtDNA PCGs, potentially due to reduced relative energy expenditure per unit mass. Furthermore, we identify several genes that undergo positive selection, possibly linked to species adaptation to specific environments. Therefore, despite purifying selection being the predominant force in the evolution of mtDNA PCGs, positive selection can also occur during the process of adaptive evolution.

P. brasiliensis and L. canadensis are two otter species, which successfully occupied semi-aquatic habitats and diverged from other Mustelidae. Herein, the full-length mitochondrial genome sequences were constructed for these two otter species for the first time. Comparative mitochondrial genome, selection pressure, and phylogenetic independent contrasts (PICs) analyses were conducted to determine the structure and evolutionary characteristics of their mitochondrial genomes. Phylogenetic analyses were also conducted to confirm these two otter species’ phylogenetic position. The results demonstrated that the mitochondrial genome structure of P. brasiliensis and L. canadensis were consistent across Mustelidae. However, selection pressure analyses demonstrated that the evolutionary rates of mitochondrial genome protein-coding genes (PCGs) ND1, ND4, and ND4L were higher in otters than in terrestrial Mustelidae, whereas the evolutionary rates of ND2, ND6, and COX1 were lower in otters. Additionally, PIC analysis demonstrated that the evolutionary rates of ND2, ND4, and ND4L markedly correlated with a niche type. Phylogenetic analysis showed that P. brasiliensis is situated at the base of the evolutionary tree of otters, and then L. canadensis diverged from it. This study suggests a divergent evolutionary pattern of Mustelidae mitochondrial genome PCGs, prompting the otters’ adaptation to semi-aquatic habitats.

Black-billed capercaillie (Tetrao parvirostris) was listed as a first-class state-protected animal because it was endangered in China (Category I). This study is the first to examine the diversity and composition of T. parvirostris gut microbiome in the wild. We collected fecal samples from five black-billed capercaillie flock roosting sites (each 20 km apart) in one day. Thirty fecal samples were sequenced with 16S rRNA gene amplicons on the Illumina HiSeq platform. This study is the first to analyze the fecal microbiome composition and diversity of black-billed capercaillie in the wild. At the phylum level, Camplyobacterota, Bacillota, Cyanobacteria, Actinomycetota, and Bacteroidota were the most abundant in the fecal microbiome of black-billed capercaillie. At the genus level, unidentified Chloroplast, Escherichia−Shigella, Faecalitalea, Bifidobacterium, and Halomonas were the dominant genera. Based on alpha and beta diversity analyses, we found no significant differences in the fecal microbiome between five flocks of black-billed capercaillie. Protein families: genetic information processing; protein families: signaling and cellular processes, carbohydrate metabolism; protein families: metabolism and energy metabolism are the main predicted functions of the black-billed capercaillie gut microbiome through the PICRUSt2 method. This study reveals the composition and structure of the fecal microbiome of the black-billed capercaillie under wild survival conditions, and this study provides scientific data for the comprehensive conservation of the black-billed capercaillie.

Building change detection (BuCD) can offer fundamental data for applications such as urban planning and identifying illegally-built new buildings. With the development of deep neural network-based approaches, BuCD using high-spatial-resolution remote sensing images (RSIs) has significantly advanced. These deep neural network-based methods, nevertheless, typically demand a considerable number of computational resources. Additionally, the accuracy of these algorithms can be improved. Hence, LightCDNet, a lightweight Siamese neural network for BuCD, is introduced in this paper. Specifically, LightCDNet comprises three components: a Siamese encoder, a multi-temporal feature fusion module (MultiTFFM), and a decoder. In the Siamese encoder, MobileNetV2 is chosen as the feature extractor to decrease computational costs. Afterward, the multi-temporal features from dual branches are independently concatenated based on the layer level. Subsequently, multiscale features computed from higher levels are up-sampled and fused with the lower-level ones. In the decoder, deconvolutional layers are adopted to gradually recover the changed buildings. The proposed network LightCDNet was assessed using two public datasets: namely, the LEVIR BuCD dataset (LEVIRCD) and the WHU BuCD dataset (WHUCD). The F1 scores on the LEVIRCD and WHUCD datasets of LightCDNet were 89.6% and 91.5%, respectively. The results of the comparative experiments demonstrate that LightCDNet outperforms several state-of-the-art methods in accuracy and efficiency.

The gastrointestinal tract of animals contains microbiota, forming a complex microecosystem. Gut microbes and their metabolites can regulate the development of host innate and adaptive immune systems. Animal immune systems maintain intestinal symbiotic microbiota homeostasis. However, relatively few studies have been published on reptiles, particularly snakes, and even fewer studies on different parts of the digestive tracts of these animals. Herein, we used 16S rRNA gene sequencing to investigate the microbial community composition and adaptability in the stomach and small and large intestines of Lycodon rufozonatus. Proteobacteria, Bacteroidetes, and Firmicutes were most abundant in the stomach; Fusobacteria in the small intestine; and Proteobacteria, Bacteroidetes, Fusobacteria, and Firmicutes in the large intestine. No dominant genus could be identified in the stomach; however, dominant genera were evident in the small and large intestines. The microbial diversity index was significantly higher in the stomach than in the small and large intestines. Moreover, the influence of the microbial community structure on function was clarified through function prediction. Collectively, the gut microbes in the different segments of the digestive tract revealed the unique features of the L. rufozonatus gut microbiome. Our results provide insights into the co-evolutionary relationship between reptile gut microbiota and their hosts.

Many mammals risk damage from virus invasion due to frequent environmental changes. The oligoadenylate synthesis (OAS) gene family, which is an important component of the immune system, provides an essential response to the antiviral activities of interferons by regulating immune signal pathways. However, little is known about the evolutionary characteristics of OASs in Laurasiatherian mammals. Here, we examined the evolution of the OAS genes in 64 mammals to explore the accompanying molecular mechanisms of the antiviral ability of Laurasiatherian mammals living in different environments. We found that OAS2 and OAS3 were found to be pseudogenes in Odontoceti species. This may be related to the fact that they live in water. Some Antilopinae, Caprinae, and Cervidae species lacked the OASL gene, which may be related to their habitats being at higher altitudes. The OASs had a high number of positive selection sites in Cetartiodactyla, which drove the expression of strong antiviral ability. The OAS gene family evolved in Laurasiatherian mammals at different rates and was highly correlated with the species’ antiviral ability. The gene evolution rate in Cetartiodactyla was significantly higher than that in the other orders. Compared to other species of the Carnivora family, the higher selection pressure on the OAS gene and the absence of positive selection sites in Canidae may be responsible for its weak resistance to rabies virus. The OAS gene family was relatively conserved during evolution. Conserved genes are able to provide better maintenance of gene function. The rate of gene evolution and the number of positively selected sites combine to influence the resistance of a species to viruses. The positive selection sites demonstrate the adaptive evolution of the OAS gene family to the environment. Adaptive evolution combined with conserved gene function improves resistance to viruses. Our findings offer insights into the molecular and functional evolution of the antiviral ability of Laurasian mammals.

Many mammals develop specific immune responses owing to the changes in their ecological niche and diet that are essential for animal survival. However, pattern recognition receptors (PRRs) serve as the first line of defense in innate immunity and generate immune responses in the host. However, the evolutionary impacts on PRR genes in Carnivora are not well studied. Herein, we explored the evolution of 946 PRR gene sequences in 43 Carnivora species to elucidate the molecular mechanisms of carnivore adaptation to complex habitats. We found that the PRRs were relatively conserved, and different gene families showed different evolutionary patterns. PRRs were highly purified based on their overall roles in Carnivora species but interspersed with positive-selection patterns during evolution. Different niche types may have jointly driven the evolution of PRR genes. In particular, the selection pressure of toll-like receptor (TLR) 10 was relaxed in seven species with pseudogenes, which may have emerged during recent evolutionary events. We speculated that a “functional compensation” mechanism may exist for genes with overlapping functions in the TLR gene family. Additionally, TLR2, TLR4, NLRC5, and DECTIN1 were subject to positive selection in semi-aquatic species, and the adaptive evolution of these genes may have been related to the adaptation to semi-aquatic environments. In summary, our findings offer valuable insights into the molecular and functional evolution of PRR genes, which are important for immune adaptations in Carnivora.

Microbial symbiotic associations may be beneficial, neutral, or harmful to the host. Symbionts exploit the host space and nutrition or use hosts as carriers to spread to other environments. In order to investigate the fecal bacterial communities of wild sika deer (Cervus nippon) and wapiti (Cervus canadensis), this study aimed to sequence and explore the composition of, and similarity between, the fecal microbiota of sika deer and wapiti using high-throughput sequencing. The composition and relative abundance of fecal microbiota, alpha diversity, and differences in beta diversity between the two species were analyzed. We found that no pathogenic bacteria were present in large quantities in the hosts. The dominant bacterial phyla found in the two deer species were similar and included Firmicutes, Bacteroidetes, Proteobacteria, and Spirochaetes. Moreover, the deer also shared similar dominant genera, including the Rikenellaceae RC9 gut group, Ruminococcaceae_UCG-010, Ruminococcaceae_UCG-005, and Bacteroides. These results demonstrate that the sika deer and wapiti share a similar fecal microbiotal structure, probably due to their common diet and living environment, but there was some evidence of a difference at the species level. These analyses provide new insights into the health status of deer populations outside protected environments and offer a scientific framework for monitoring the health conditions of sika deer and wapiti.

The Gobiidae family occupy one of the most diverse habitat ranges of all fishes. One key reason for their successful colonization of different habitats is their ability to adapt to different energy demands. This energy requirement is related to the ability of mitochondria in cells to generate energy via oxidative phosphorylation (OXPHOS). Here, we assembled three complete mitochondrial genomes of Rhinogobius shennongensis, Rhinogobius wuyanlingensis, and Chaenogobius annularis. These mitogenomes are circular and include 13 protein-coding genes (PCGs), two rRNAs, 22 tRNAs, and one non-coding control region (CR). We used comparative mitochondrial DNA (mtDNA) genome and selection pressure analyses to explore the structure and evolutionary rates of Gobiidae mitogenomics in different environments. The CmC model showed that the ω ratios of all mtDNA PCGs were adenosine triphosphate 8 (atp8) was faster in Gobiidae than in other mitochondrial DNA PCGs. We also found evidence of positive selection for several sites of NADH dehydrogenase (nd) 6 and atp8 genes. Thus, divergent mechanisms appear to underlie the evolution of mtDNA PCGs, which might explain the ability of Gobiidae to adapt to diverse environments. Our study provides new insights on the adaptive evolution of Gobiidae mtDNA genome and molecular mechanisms of OXPHOS.

Microtubule actin crosslinking factor 1 (MACF1) is a large crosslinker that contributes to cell integrity and cell differentiation. Recent studies show that MACF1 is involved in multiple cellular functions such as neuron development and epidermal migration, and is the molecular basis for many degenerative diseases. MACF1 is highly abundant in bones, especially in mesenchymal stem cells; however, its regulatory role is still less understood in bone formation and degenerative bone diseases. In this study, we found MACF1 expression in mesenchymal stem cells (MSCs) of osteoporotic bone specimens was significantly lower. By conditional gene targeting to delete the mesenchymal Macf1 gene in mice, we observed in MSCs decreased osteogenic differentiation capability. During early stage bone development, the MACF1 conditional knockout (cKO) mice exhibit significant ossification retardation in skull and hindlimb, and by adulthood, mesenchymal loss of MACF1 attenuated bone mass, bone microarchitecture, and bone formation capability significantly. Further, we showed that MACF1 interacts directly with SMAD family member 7 (SMAD7) and facilitates SMAD7 nuclear translocation to initiate downstream osteogenic pathways. Hopefully these findings will expand the biological scope of the MACF1 gene, and provide an experimental basis for targeting MACF1 in degenerative bone diseases such as osteoporosis.