Sign Language Translation (SLT) is a challenging task due to its cross-domain nature, involving the translation of visual-gestural language to text. Many previous methods employ an intermediate representation, i.e., gloss sequences, to facilitate SLT, thus transforming it into a two-stage task of sign language recognition (SLR) followed by sign language translation (SLT). However, the scarcity of gloss-annotated sign language data, combined with the information bottleneck in the mid-level gloss representation, has hindered the further development of the SLT task. To address this challenge, we propose a novel Gloss-Free SLT based on Visual-Language Pretraining (GFSLT-VLP), which improves SLT by inheriting language-oriented prior knowledge from pre-trained models, without any gloss annotation assistance. Our approach involves two stages: (i) integrating Contrastive Language-Image Pre-training (CLIP) with masked self-supervised learning to create pre-tasks that bridge the semantic gap between visual and textual representations and restore masked sentences, and (ii) constructing an end-to-end architecture with an encoder-decoder-like structure that inherits the parameters of the pre-trained Visual Encoder and Text Decoder from the first stage. The seamless combination of these novel designs forms a robust sign language representation and significantly improves gloss-free sign language translation. In particular, we have achieved unprecedented improvements in terms of BLEU-4 score on the PHOENIX14T dataset (≥+5) and the CSL-Daily dataset (≥+3) compared to state-of-the-art gloss-free SLT methods. Furthermore, our approach also achieves competitive results on the PHOENIX14T dataset when compared with most of the gloss-based methods 1 .
Background: Function genomic studies will generally result in lists of genes that may provide clues for exploring biological questions and discovering unanticipated functions, based on differential gene expression analysis, differential epigenomic analysis or co-expression network analysis. While tools have been developed to identify biological functions that are enriched in the genes sets, there remains a need for comprehensive tools that identify functional enrichment of genes for both model and non-model species from a different function classification perspective. Results: We developed AllEnricher, a tool that calculates gene set function enrichment, with user-defined updatable libraries backing up for both model and non-model species as well as providing comprehensive functional interpretation from multiple dimensions, including GO, KEGG, Reactome, DO and DisGeNET. Conclusions: AllEnricher incorporates up to date information from different public resources and provides a comprehensive resolution for biologists to make sense out of specific gene sets, making it an advanced open-source tool for gene set function analysis.
The reed vole (Microtus fortis) is an important rodent model for studying unique biological traits, such as its natural resistance to Schistosoma japonicum. To facilitate the genetic study of these phenotypes, we have produced the first high-quality, chromosome-level genome assembly for this species. The genome was assembled using PacBio HiFi long-read sequencing and scaffolded to the chromosome level with Hi-C data. The final 2.29 Gb assembly exhibits excellent continuity (contig N50 = 68.89 Mb; scaffold N50 = 91.23 Mb), with 97.7% of the sequence anchored into 26 pseudomolecules, consistent with the species’ karyotype. Genome completeness was estimated at 96.3% via BUSCO analysis (glires_odb10). The annotation includes 23,678 protein-coding genes, with 97.5% assigned a putative function. This publicly available, high-quality genomic resource will be invaluable for future research, providing the necessary foundation to explore the genetic mechanisms behind the unique adaptations of M. fortis, including its innate immunity, digestive physiology, and disease models. The assembly will also serve as a key reference for comparative genomics, enriching our understanding of rodent evolution.
Human conservation efforts have pros and cons for endangered animals. Metagenomic analysis of the gut microbiome provides valuable insights and tools for wildlife conservation. However, most studies only focus on changes in microbial species and potential functional genes, leaving out the assembly mechanisms and metabolic activities of wildlife gut microbiome. Herein, we investigated the gut microbiome of golden snub-nosed monkeys under varied conservation strategies (wild, captive, and food provision) using metagenome and metabolome analysis. We found that the conserved monkeys had more non-redundant genes, distinct community diversities, and less stable gut microbiota. Source-tracking analysis showed captive group has fewer wild-sourced microbes than food provision group. The captive group had more habitat specialists with narrower niche breadth compared to the provision/wild counterparts. Besides, captive group's microbiota governed by deterministic processes more than provisioned/wild groups. Procrustes analysis revealed changes in antibiotic resistance genes and virulence factors linked to mobile genetic elements in conserved monkeys. Notably, gut microbiota and metabolomic dataset had significant co-variation and clear separation among the wild, captive and food provision groups. Weighted correlation network analysis identified co-varying modules of microbiota and metabolites with opposite variation trends between the wild and conserved animals, specifically linked to co-metabolism pathways associated with essential amino acids. This study provides new insights into the structural features, assembly mechanisms, community functions, and microbiota-metabolome associations of wildlife gut microbiome under different conservation strategies and have significant implications for advancing sustainable conservation practices.
Human conservation efforts have pros and cons for endangered animals. Metagenomic analysis of the gut microbiome provides valuable insights and tools for wildlife conservation. However, most studies only focus on changes in microbial species and potential functional genes, leaving out the assembly mechanisms and metabolic activities of wildlife gut microbiome. Herein, we investigated the gut microbiome of golden snub-nosed monkeys under varied conservation strategies (wild, captive, and food provision) using metagenome and metabolome analysis. We found that the conserved monkeys had more non-redundant genes, distinct community diversities, and less stable gut microbiota. Source-tracking analysis showed captive group has fewer wild-sourced microbes than food provision group. The captive group had more habitat specialists with narrower niche breadth compared to the provision/wild counterparts. Besides, captive group’s microbiota governed by deterministic processes more than provisioned/wild groups. Procrustes analysis revealed changes in antibiotic resistance genes and virulence factors linked to mobile genetic elements in conserved monkeys. Notably, gut microbiota and metabolomic dataset had significant co-variation and clear separation among the wild, captive and food provision groups. Weighted correlation network analysis identified co-varying modules of microbiota and metabolites with opposite variation trends between the wild and conserved animals, specifically linked to co-metabolism pathways associated with essential amino acids. This study provides new insights into the structural features, assembly mechanisms, community functions, and microbiota-metabolome associations of wildlife gut microbiome under different conservation strategies and have significant implications for advancing sustainable conservation practices.