Biology subjects

Jiang, Y.

Publications and source records attributed to Jiang, Y..

At least 127 records · Page 7Linked to original sources

bioRxiv2025

Enhancing the iNKT cell immunotherapy platform by combining optimised CAR endodomains with novel iNKT engagers

iNKT cells are emerging as a highly promising immunotherapy platform for the treatment of cancer. To maximise the anti-cancer activity of CAR-iNKT against the blood cancer multiple myeloma we investigated optimal CAR designs and their combination with novel iNKT-specific engagers. We find that amongst five different CAR endodomains, underpinned by increased avidity and a cross talk between Plexin D1 on CAR-iNKT and Semaphorin 4A on myeloma cells, BCMA CD28z CAR-iNKT exert the highest anti-myeloma activity. Notably, CD28z CAR-iNKT outperform their CAR-T counterparts. To expand the anti-myeloma potential of CAR-iNKT, we designed and validated a high efficacy BCMA iNKT-specific engager which exerts significant anti-myeloma activity in conjunction with adoptively transferred iNKT cells. Finally, combined, dual target therapy with FCRL5 CAR-iNKT and BCMA iNKT engagers outperforms FCRL5 CAR-iNKT and limits immune escape of FCRL5-negative myeloma. Thus, optimised iNKT-based, dual-target, dual-modality immunotherapy has enhanced anti-tumor activity against multiple myeloma and potentially other malignancies. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=159 SRC="FIGDIR/small/683869v1_ufig1.gif" ALT="Figure 1"> View larger version (33K): org.highwire.dtl.DTLVardef@11e843corg.highwire.dtl.DTLVardef@7bb408org.highwire.dtl.DTLVardef@1687b84org.highwire.dtl.DTLVardef@8542ea_HPS_FORMAT_FIGEXP M_FIG C_FIG

immunology↗

bioRxiv2025

Psychological stress and social support are associated with opposing single-cell pro-inflammatory gene regulatory mechanisms in adults

Psychological stress is linked to elevated markers of chronic inflammation, whereas social support is associated with lower levels; yet, the molecular mechanisms mediating these effects are poorly understood. We investigated gene regulatory variation in peripheral blood mononuclear cells (PBMCs) from 165 self-reported African American adults (aged 50-89 years) using single-cell RNA sequencing (scRNA-seq) and single-cell chromatin accessibility (scATAC-seq). Self-reported psychological stress and social support were associated with differential expression of 1,956 and 1,296 genes, respectively (10% FDR), primarily in CD4+ T cells and monocytes. Interferon signaling genes showed high expression in individuals with high psychological stress and low expression in those with high social support; this pattern mirrored gene expression in individuals with elevated circulating inflammatory markers (IFN-{gamma}, TNF-, IL-6). Genome-wide transcription factor (TF) motif analysis identified stress- and social support-associated changes in motif activity for 70 and 116 TFs, respectively, with 87 motifs enriched near differentially expressed genes. In CD4+ T cells, high psychological stress corresponded to increased IRF and STAT TF motif activity (interferon pathway), while social support was associated with reduced activity and expression in these pathways. We used an immune challenge paradigm (i.e., LPS stimulation), which confirmed the biological pathways of these gene regulatory effects. Our results demonstrate that psychological stress and social support modulate immune gene regulation at the single-cell level, revealing mechanistic links between psychosocial factors and inflammation, and suggesting that social support may promote immunological health.

genomics↗

bioRxiv2025

ctsGRN: Inferring cell type-specific gene regulatory networks in Arabidopsis

Gene regulatory networks (GRNs) govern life processes. And exploring GRNs is of great value for understanding biological functions and developing species resources. At present, studies on plant GRNs are limited, mostly focusing on the tissue level, which hinders the discovery of cell type-specific regulatory networks and mechanisms. To address this, we have proposed a GRN research framework named ctsGRN (CellTypeSpecGRN) for conducting cell-specific GRN research at the cell type resolution. In this framework, we employed three representative algorithms from statistics (e.g., correlation analysis), traditional machine learning (e.g., random forest) and deep learning (e.g., neural networks) to construct GRNs for different cell types in Arabidopsis roots. Among these, traditional machine learning methods showed the best performance. Experimental validation through yeast one-hybrid (Y1H) assays and ChIP-seq data confirmed the high reliability and predictive power of these inferred GRNs. By analyzing these networks, we mined novel cell type-specific regulatory relationships, identified core transcription factor (TF) AT5G08790 involved in root development, discovered 249 cell-type-specific hub TFs, and 136 key functional modules were found to be enriched in biological processes. Additionally, we revealed significant heterogeneity of TF-target gene interactions across various cell types. This comprehensive analysis offers a detailed overview of TF-mediated cell type-specific transcriptional regulation in Arabidopsis roots, providing new insights into the molecular mechanisms of root development and establishing a guideline framework for GRN research at the cell-type level in plants.

bioinformatics↗

bioRxiv2025

RUNX1T1-HDAC Reprogramming of the HOX Code Signaling Drives a Targetable Pan-Cancer Lineage Plasticity

Cancer remains a leading cause of death worldwide, with lineage plasticity emerging as a hallmark that drives therapy resistance and tumor progression by enabling cancer cells to alter identity and evade targeted therapies. Although genomic and transcriptomic aberrations correlate with lineage plasticity, the absence of pan-cancer markers to rapidly identify plastic subtypes has limited predictive utility. Homeobox (HOX) genes encode transcription factors that define tissue identity through distinct expression patterns, or HOX codes, in specific lineages. By analyzing multi-omics data, including 39 HOX genes from over 80,000 RNA-seq samples across 114 cancer types, we discovered that HOX code expression robustly represents cancer cell lineages and reveals multiple previously unrecognized lineage-plastic subtypes in prostate cancer, lung cancer, and acute myeloid leukemia (AML), each displaying altered HOX patterns compared to non-plastic subtypes. Differential analysis further identified RUNX1T1 as a novel and consistent marker of plasticity, elevated across all three cancer types and correlating with HOX code and lineage-plastic marker genes. We validated these correlations in bulk and single-cell RNA-seq from extensive preclinical and clinical cohorts and provided direct functional evidence that RUNX1T1 is required for lineage-plastic programs in prostate cancer. Using AI-based modeling, we identified NCOR/HDAC as RUNX1T1 binding partners forming a co-repressor complex that regulates HOX codes and plasticity. Finally, pharmacologic HDAC inhibition selectively suppressed the growth of plastic cells, revealing a novel therapeutic vulnerability. These findings establish ectopic RUNX1T1 as a pan-cancer biomarker and critical mediator of lineage plasticity and identify the RUNX1T1-HDAC complex as a druggable target.

cancer biology↗

bioRxiv2025

Proteome Landscapes Decode Organelle Vulnerabilities in cortical and dopaminergic-like induced neurons Across Lysosomal Storage Disorders

Lysosomes maintain cellular homeostasis by degrading proteins delivered via endocytosis and autophagy, and by recycling building blocks for organelle biogenesis. Lysosomal Storage Disorders (LSDs) comprise a group of diseases affecting diverse lysosomal functions. To facilitate molecular phenotyping across diverse LSD gene classes, we are developing a library of human embryonic stem cells engineered to lack individual LSD genes as a resource for the field. Here, we report our initial stem cell toolkit lacking one of 23 LSD genes, including the majority of genes associated with sphingolipidoses and neuronal ceroid lipofuscinoses, and its use in the generation of a proteomic resource for induced cortical-like and midbrain dopaminergic-like neurons In-depth abundance and correlation profiling across organelles and sub-organelle components revealed potential vulnerabilities that reflect distinct patterns of proteome alterations across both genotypes and neuronal cell types. We characterize alterations in the mitochondrial proteome associated with GBA1 and ASAH1 deficiency and identify synaptic and mitochondrial defects in ASAH1-/- induced neurons that correlate with defects in neuronal firing rates. Moreover, we developed an informatic pipeline for proteome-wide identification of individual protein interactions and protein complexes that may be disrupted as a result of LSD gene deficiency. Finally, we visualized structural alterations of ASAH1-deficient endolysosomes in situ using cryo-electron tomography, revealing swollen organelles that were largely devoid of dense internal membranes characteristic of wild-type cells, but containing numerous intralumenal vesicle compartments. This toolkit and associated proteomic landscapes provide a resource for defining molecular signatures associated with LSD gene dysfunction and organelle vulnerability.

cell biology↗

bioRxiv2025

Epigenetic characterization of pseudogenes across human tissues

Pseudogenes have long been considered non-functional remnants of genome evolution. However, compared to other noncoding elements, their epigenetic regulation remains poorly understood, largely due to their sequence similarity to parent genes and the lack of systematic functional analyses. Here, we leverage matched transcriptomic and epigenomic data from the ENCODE EN-TEx project, spanning [~]30 human tissues, to comprehensively profile pseudogenes and compare them to protein-coding transcripts and lncRNAs. Our analyses reveal epigenetic differences among pseudogene biotypes. Even when transcribed, processed pseudogenes--unlike protein-coding transcripts and unprocessed pseudogenes--lack canonically active histone marks and open chromatin signatures at their promoters. Instead, these regions are frequently associated with retrotransposon sequences, enriched for YY1-binding sites, and often engage distal enhancer-like elements through long-range genomic contacts. These findings suggest that processed pseudogenes may adopt regulatory mechanisms akin to transposable elements such as LINEs, relying on YY1-mediated enhancer-promoter looping or other distal interactions to facilitate transcription. By publicly releasing our catalog of pseudogene promoter regions, we provide a comprehensive resource for investigating the epigenetic and functional roles of pseudogenes in the human genome.

bioinformatics↗

bioRxiv2025

Algorithm-Powered Analyzer for Continuous Electrochemistry: A Toolkit for Real-Time Electrochemical Data Analysis

Real-time biosensors offer significant potential for continuous monitoring of biomolecules. However, their practical application and further development face challenges on data analysis, including poor signal-to-noise ratio when effective sensing area decreases due to signal degradation by biofouling, time-consuming and subjective process of manual or semi-automated peak identification, and inconsistencies in data interpretation, thereby complicating reproducibility and cross-comparison of biosensing results. In this study, we introduce the Algorithm-Powered Analyzer for Continuous Electrochemistry (A-PACE), an open-source toolkit providing streamlined and automated data analysis protocol optimized for real-time electrochemical data analysis. A-PACE comprises three modules: (1) Change point detection for automated identification of peak regions; (2) Baseline fitting with multiple algorithms to handle diverse electrochemical signals and baseline screening to eliminate unreasonable fits; (3) Large dataset input with user-friendly interface for data processing, exporting and visualization. To ensure optimal performance, we curated a training set of 2000 electrochemical curves from an extensive electrochemical dataset (>100,000 curves) collected over the past five years under varied electrochemical measurement conditions. These curves were analyzed using 2046 algorithm sets to identify a default algorithm set, compared to existing tools, demonstrating its capability in peak detection and baseline fitting across a broad spectrum of electrochemical data. Case studies reveal the A-PACE can analyze month-long in vitro serum data and week-long in vivo intravenous blood data, extending the operational lifespan of real-time sensors. This cross-platform compatible toolkit supports both real-time and post-processing analysis, reducing the processing time from minute level per signal by human labeling to second level by A-PACE and subjectivity associated with electrochemical signal processing. By providing this solution for continuous electrochemical data analysis, A-PACE enhances biosensors applications in medical diagnostics and continuous monitoring with high throughput analysis.

bioengineering↗

bioRxiv2025

Illuminating the Virosphere's Dark Matter using Hierarchical Deep Learning

Systematic discovery of novel viruses is essential for pandemic preparedness, understanding tumor-associated viruses, developing viral delivery systems, and advancing biomedical applications. Yet, the majority of sequences in metagenomic datasets lack close relatives in existing references, representing a vast viral "dark matter" whose biology and evolution remain largely unknown. The central task is threefold: 1) to determine whether a genome is viral or non-viral, 2) to correctly assign viral genomes to known lineages when possible, and, critically, 3) to recognize when no existing lineage applies and thereby identify candidates for entirely novel viral groups. Existing approaches, which depend on sequence homology or narrow markers, struggle to capture this uncharted viral space. Here we present DeepVirus, a hierarchical transformer-based framework that models viral genomes as structured sequences of protein-coding genes. By combining protein-level embeddings from a foundation model with genome-aware representations, DeepVirus not only achieves accurate classification across deep taxonomic hierarchies, but also extends beyond conventional classification to detect and organize candidate novel viral lineages through open-set recognition. Applied to large-scale metagenomic resources, DeepVirus uncovered extensive viral diversity, including previously uncharacterized RNA-dependent RNA polymerases (RdRps), thereby expanding the known evolutionary space of RNA viruses. DeepVirus integrates deep learning with genome-aware open-set discovery to illuminate viral dark matter, providing a foundation for systematic viral taxonomy and advancing exploration of the global virosphere, with broad implications for safeguarding human health.

bioinformatics↗

bioRxiv2025

Adaptation of Perceived Animacy from Biological Motion: Evidence for a "Life Motion Detector"

Humans can readily perceive animacy from biological motion (BM) - the distinctive movement patterns of living entities. However, how the human brain extracts animacy information from BM remains largely unclear. The current study investigated this issue using visual adaptation, a non-invasive tool for revealing neural mechanisms underlying the selective processing of specific properties. Results showed that prolonged exposure to intact human walkers, compared to non-BM adaptors, biased the perception of subsequent walking stimuli towards less animate. This adaptation aftereffect persisted following adaptation to local foot motions carrying diagnostic kinematic cues, but was absent after exposure to static body forms, indicating the involvement of neuronal populations encoding animacy from BM based on motion signals. Moreover, adapting to pigeon movements also biased animacy perception for human motions, revealing a shared mechanism for encoding animacy from cross-species BM signals. These results support the existence of an inherent "life detection" system in the human brain, attuned to local foot motions and kinematic features prevalent in vertebrate movements, which may lay a foundation for perceiving life from motion across terrestrial species. Public significance statementsDiscerning whether a moving entity is animate is crucial for survival and for generating appropriate responses. This study explored how the human brain perceives animacy based on a prevalent movement pattern among vertebrates, namely, walking. Leveraging a non-invasive visual adaptation paradigm, five experiments demonstrated that pre-exposure to intact human walking, feet-only walking, and even intact pigeon walking motion sequences - but not static human forms - elicited significant adaptation aftereffects in animacy perception of human walkers. These findings support the existence of neural substrates encoding perceived animacy based on biological motion signals, particularly those carried by local foot motions and cross-species kinematic features. Such mechanisms may be integral to an inherent life detection system in the human brain.

neuroscience↗

bioRxiv2025

Comprehensive analysis of pseudogene expression in human and macaque brains compared with other tissues

Although gene expression in the brain has been extensively investigated and compared to other tissues, the activity of pseudogenes has not been comprehensively surveyed. Here, leveraging large-scale RNA-seq data, we construct consistent pseudogene expression profiles in human and macaque brains and compare them to 29 other tissues. We further annotate pseudogenes with potential cellular roles based on co-clustering them with protein-coding genes. Notably, the majority of the expressed pseudogenes show elevated expression in the brain relative to other tissues, and these pseudogenes show broad and consistent expression patterns across brain subregions. Furthermore, spatiotemporal analyses reveal that pseudogenes in different brain subregions have greatly varying temporal trajectories (e.g., increasing vs. decreasing), in contrast to protein-coding genes that tend to be more uniform. Finally, we identify a set of pseudogenes exhibiting significant changes in neuropsychiatric disorders, some of which overlap with known brain eGenes (genes whose expression is associated with expression quantitative trait loci, eQTLs), as well as with genes implicated by genome-wide association studies (GWAS). Together, our study provides a public resource of pseudogene expression in human and macaque brains (in comparison to other tissues) and highlights aspects of pseudogene expression in brain development and pathogenesis.

bioinformatics↗

bioRxiv2025

First Whole-Genome Chikungunya Virus Sequence Detected in Mosquitoes during the 2025 Foshan Outbreak: Evidence of Field Vector Infection and Transmission Potential in China

BackgroundSince July 2025, an outbreak of mosquito-borne chikungunya fever occurred in Foshan City, Guangdong Province, China. This was the second outbreak in China following the one that occurred in Dongguan City, Guangdong Province, in 2010. Moreover, the intensity of this outbreak was significantly greater than that of the previous one. Updates to 23 August, more than 10,000 human cases had been reported. Here, we present the first full genome sequence of the chikungunya virus (CHIKV) derived from field-trapped mosquitoes during the outbreak. MethodsAdult Aedes albopictus were BG-trap captured from residences and parklands in three hotspot towns with high density of confirmed human cases. Mosquitoes were morphologically identified and pooled by species, sex and environment types. RNA was extracted, screened by CHIKV RT-qPCR, then positive pools underwent sanger and whole-genome sequencing for complete sequences. Lineage and mutational profiles were inferred by maximum likelihood phylogenetic and comparison against human and mosquito genomes. The distribution of amino acid site mutations in different protein coding regions was also analyzed. ResultThrough 11 days of collection using 10 BG-traps, 2,803 mosquitoes were captured. 1569 (55.97%) female Ae. albopictus were divided into 77 pools and 9.09% (7/77) of the pools tested positive for CHIKV. The local Ae. albopictus minimum infection rate (MIR, per 1000 females) was 4.46, while the MIR for residences in Lecong Town was the highest at 9.17. The MIR for parklands was slight higher than for residences (4.60 vs. 4.30). All the 5 Ae. albopictus-derived complete CHIKV genome clustered within ECSA-Indian Ocean lineage genotype, closely related to human-derived genomes on 2025 Reunion Island. Amino-acid mutations E1-A226V/E2-L210Q were detected in the strains, which enhanced adaptability to Ae. albopictus and increased the transmission capacity. Novel mutation observed on E1 and E2 were totally consist to the patient-derived CHIKV in 2025 Reunion Island. ConclusionsIt was the first mosquito-derived CHIKV whole-genome during the 2025 Foshan outbreak, filling a critical gap between human case and entomological surveillance. Ae. albopictus was confirmed as the primary vector during the outbreak. The current outbreak CHIKV strain with particular amino-acid mutations had adapted to Ae. albopictus transmission. Compared to previous Chikungunya outbreaks over the past decade, the Foshan outbreak occurred earlier (early July), in a larger urban area (with a population of over 9.5 million), and with abundant breeding sites for the vector mosquito Ae. albopictus. However, the outbreak was quickly brought under control, with daily case numbers consistently decreasing, which is closely linked to the strong vector control measures implemented by the Chinese government in the early stages of the outbreak. Moreover, this event once again underscores the necessity of early monitoring of vector mosquitoes and the importance of implementing highly effective vector intervention measures as soon as possible after an outbreak occurs.

microbiology↗

bioRxiv2025

Augment rice ANNEXIN expression to counter planthopper NlAnnexin-like5 as an anti-virulence strategy against a major crop pest

The brown planthopper (BPH) is the most devastating insect pest in rice, posing a serious threat to global rice production. One attractive control strategy would be based on the understanding of the virulence mechanisms of BPH at the molecular level and then designing targeted methods to neutralize such mechanisms. Salivary proteins of BPH are important players in mediating rice-BPH interactions. Here, we describe a pivotal role of a watery saliva protein, NlAnnexin-like5 (NlANX5), in the rice-BPH interaction. RNA interference (RNAi) of NlANX5 greatly compromised BPH feeding performance and survival rate on rice plants. NlANX5-RNAi BPH triggered a rapid calcium ion influx in rice cells. The feeding and survival defects of NlANX5-RNAi BPH can be restored in NlANX5-expressing transgenic rice plants. NlANX5 targets rice annexin (OsANN) proteins, including OsANN2 and OsANN8. Further analysis with NlANX5 and OsANN2 as well as OsANN8 showed that NlANX5 displaces OsANN2 and OsANN8 from the rice cell membrane. The osann2 osann8 mutant rice plants are hypersusceptible to BPH infestation. In contrast, enhanced expression of OsANN2 and OsANN8 gene resulted in robust rice resistance against BPH. This study highlights a successful example of identifying and augmenting the expression of the host targets of a major BPH virulence effector as a promising anti-virulence strategy against an important crop pest. Significance StatementInsect pests pose a serious biotic threat to crop production worldwide. Understanding how insect pests attack plants could inspire innovative pest control measures to enhance global food security. The brown planthopper (BPH; Nilaparvata lugens St[a]l) is the most devastative insect pest in rice. In this study, we discovered that BPH secretes a salivary protein, NlAnnexin-like5 (NlANX5), to target rice host annexins that are associated with calcium fluxes and activation of multiple rice defense pathways. This knowledge led to employing enhanced expression of rice annexin-encoding genes to successfully defeat the virulence function of NlANX5. Results have significant implications in the development of anti-virulence breeding strategies against BPH.

plant biology↗

bioRxiv2025

SARS-CoV-2 S protein activates the HIV latent reservoir through the mTOR pathway

During the global COVID-19 pandemic, mRNA vaccines using the S protein as antigen were widely used.Vaccine-induced S proteins can persist in vivo for weeks, triggering low-level immune activation. HIV latent reservoir maintenance is a major challenge for ART therapy, especially when immune pressure is waning. This then raises critical questions for HIV-infected patients: does prolonged exposure to S proteins affect HIV latent reservoir stability? Recent studies have pointed out that S proteins may activate the mTOR signaling pathway, which in turn affects the immune response and metabolic processes of cells. And the mTOR pathway is closely related to the maintenance and activation of HIV latent reservoir. However, how S proteins affect the HIV latent reservoir and the mechanism of activation are unclear. To explore the mechanism of how SARS-CoV-2 S proteins regulate the HIV latent reservoir and to explore whether S proteins regulate the HIV latent reservoir through the mTOR pathway, we constructed an in vitro HIV latent reservoir model for our experiments.To evaluate the potential role of S protein in HIV latent reservoir activation, relevant markers of HIV latent reservoir activation were detected using ELISA, flow cytometry, and RT-qPCR; and the relationship between S protein and mTOR was also detected by WB, CO-IP, and IFC.It was found that S proteins activated the HIV latent reservoir while increasing mTOR expression. It was further observed that mTOR inhibitors significantly inhibited S protein-induced activation of the HIV latent reservoir, and mTOR activators reversed the inhibitory effect of mTOR inhibitors on HIV latent reservoir activation. In summary, we found that S proteins activated the HIV latent reservoir through the mTOR pathway. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=170 SRC="FIGDIR/small/670887v1_ufig1.gif" ALT="Figure 1"> View larger version (48K): org.highwire.dtl.DTLVardef@f6d87aorg.highwire.dtl.DTLVardef@1fe4cc0org.highwire.dtl.DTLVardef@60eefeorg.highwire.dtl.DTLVardef@1ea9a34_HPS_FORMAT_FIGEXP M_FIG S protein interacts with mTOR and activates the mTOR-p-p70S6K-pS6 pathway, which promotes HIV transcription C_FIG

immunology↗

bioRxiv2025

PrimateFace: A Machine Learning Resource for Automated Face Analysis in Human and Non-human Primates

Machine learning has revolutionized human face analysis, but equivalent tools for non-human primates remain limited and species-specific, hindering progress in neuroscience, anthropology, and conservation. Here, we present PrimateFace, a comprehensive, cross-species platform for primate facial analysis comprising a systematically curated dataset of 260,000+ images spanning over 60 genera, including a genus-balanced subset of 60,000 images, annotated with bounding boxes and facial landmark configurations. Face detection and facial landmark estimation models trained on PrimateFace achieve high cross-species performance, from tarsiers to gorillas, achieving performance comparable to baseline models trained exclusively on human data (0.34 vs. 0.39 mAP for face detection; 0.061 vs. 0.053 normalized landmark error), demonstrating the generalization benefits of cross-species training. PrimateFace enables diverse downstream applications including individual recognition, gaze analysis, and automated extraction of stereotyped (e.g., lip-smacking) and subtle (e.g., soft left turn) facial movements. PrimateFace provides a standardized platform for facial phenotyping across the primate order, empowering data-driven studies that advance the health and well-being of human and non-human primates. All models, notebooks, and data can be found at github.com/KordingLab/PrimateFace.

animal behavior and cognition↗

bioRxiv2025

Helicobacter pylori genome aggregation database reveals complex evolutionary forces shaping its genomic landscape and clinical impact in East Asia

We assembled HPgnomAD, a comprehensive global genome aggregation database of Helicobacter pylori (H. pylori) featuring 7,544 high-quality genomes from 278 populations, along with the first species-wide haplotype reference panel for genotype imputation. The panel provides high accuracy across diverse H. pylori populations, including those from low-coverage data, and is openly accessible with integrated analysis tools (https://www.hpgnomad.top/). Variant discovery revealed 1.82 million SNPs and 0.65 million InDels, with African strains exhibiting the most tremendous diversity and East Asian strains showing unexpectedly high novel variation. Fine-scale analysis of the hpEAsia lineage revealed six new sublineages shaped by altitude-, latitude-related divergence and region-specific admixture. Phylogenomic dating revealed two divergence waves in East Asia, paralleling Upper Paleolithic settlement and Neolithic human expansions, with highland-lowland separation at [~]18.8 kya, followed by the formation of complex geography-related population substructures. Genome-wide scans revealed adaptive loci related to metal acquisition, nitrogen metabolism, surface adhesion, and membrane transport, including altitude-associated highly differentiated variants linked to antibiotic resistance and gastric disease. By integrating genomic, evolutionary, environmental, and clinical data, HPgnomAD offers a framework for understanding H. pylori evolution, host-pathogen adaptation, and precision medicine worldwide.

evolutionary biology↗

bioRxiv2025

Protein Function Prediction via Contig-Aware Multi-Level Feature Integration

Proteins play a central role in biological processes, and accurately predicting their functions is crucial for biomedical research. While computational methods have advanced significantly, most approaches rely solely on sequence or structure, neglecting critical inter-protein relationships, such as the topological arrangement of coding sequences (CDSs) within contigs. To address this gap, we propose CAML, a novel deep learning model that integrates intra-protein features including sequence and predicted structure with inter-protein features capturing functional linkages among CDSs in contigs. Specifically, CAML employs a Graph Isomorphism Network (GIN) to extract structural features from predicted protein contact graphs and ESM-2 for sequence embeddings. Additionally, it leverages kmer frequencies and a Bidirectional Long Short-Term Memory (BiLSTM) network to model functional relationships among colocalized CDSs within contigs, capturing operon-like associations. Extensive experiments demonstrate that CAML outperforms the state-of-the-art methods in accuracy, precision, recall and F1-score, achieving improvements of 11.24%, 12.43%, 13.59%, and 13.30%, respectively over the second-best model. Ablation studies further confirm the critical contribution of CAMLs multi-level biological feature integration in enhancing functional annotation accuracy. Our study demonstrates the importance of integrating structural, sequential, and CDSs topological features for accurate protein function prediction, providing a robust computational framework for genomics research.

bioinformatics↗

bioRxiv2025

Cost-effective genomic prediction for fertility traits: A comparison of machine learning and conventional models using low-coverage sequencing in Holstein heifers

Fertility is one of the major factors affecting the efficiency of dairy herd, and genomic selection (GS) on milk yield, while ignoring fertility, has resulted in a decline in heifer fertility. Consider the cost of breeding, it is important to enhance the accuracy of GS with cost-effective way for fertility traits. This study investigates the genomic prediction (GP) of fertility traits in Holstein heifers using both machine learning (ML) and conventional methods, based on data from SNP arrays and low-coverage sequencing data. In this study, we collected 45,320 Holstein heifers with phenotype and pedigree records, from which we generated genomic data for 3,683 Holstein heifers using lcGWS. We first estimated the heritability for age at first service (AFS), gestation length (GLh) and age at first calving (AFC). We then compared the prediction performance of ML methods, kernel ridge regression (KRR), support vector regression, and random forest regression, with GBLUP, ssGBLUP and BayesR3 regarding GP accuracy and unbiasedness. Inputs for ML includes genomic relationship matrices (GRM), principal components, and SNPs. The results revealed that the heritability for the three fertility traits ranged from 0.09 to 0.48. Prediction accuracy from imputed low-coverage sequencing data was comparable to that from standard SNP chips. When both pedigree and genotypic data were used for GS, ssGBLUP yielded the highest prediction accuracy for AFS and GLh. Crucially, using only genomic data, KRR_GRM improved GP accuracy by up to 28.57% compared to GBLUP and by up to 9.46% compared to BayesR3. Our results highlight the effectiveness of low-coverage sequencing data in breeding applications and the MLs potential to enhance GP accuracy for fertility traits, offering practical insights for dairy breeding programs. ImplicationsTo improve the accuracy of GS with cost-effective way for fertility traits, this study confirms that low-coverage sequencing offers both accuracy and cost-effectiveness for fertility in Holstein heifers. The research provides a decision-making framework for breeding workers: the ssGBLUP model is optimal when combining pedigree data, while machine learning methods are superior with only genomic data. This study offers a practical tool for achieving efficient and economical genetic improvement in dairy cattle.

genomics↗

bioRxiv2025

Genomic Regions and Variant Analysis Reveal Candidate Genes Associated with Age at First Service in Chinese Holstein Heifers

Understanding the genetic basis and identifying quantitative trait loci (QTL) for age at first service (AFS) is essential for improving reproductive efficiency and reducing economic costs in heifers rearing. We conducted a genome-wide association study (GWAS) for AFS using the genomic estimated breeding values (GEBVs) of 3,686 Chinese Holstein heifers, genotyped with 7,964,217 single nucleotide polymorphisms (SNPs) after quality control. Significant SNPs and candidate genes were identified and investigated through colocalization analysis of GWAS, expression QTL (eQTL), and splicing QTL (sQTL), and expression analysis in blood RNA sequencing (RNA-seq). The heritability estimate for AFS was moderate at 0.276 {+/-}0.025. Three QTL regions associated with AFS were identified: Region 1 on BTA6 (43688837-45007127 bp, 1.3 Mb with three 200kb windows), explaining 0.47% of genetic variance; Region 2 on BTA6 (62770788-62967316 bp) explaining 0.27% of the genetic variance; and Region 3 on BTA18 (6299721-6498934 bp) explaining 0.17% of the genetic variance. Combining GWAS and colocalization analysis, we identified ANAPC4, RBPJ, SEPSECS, SLC34A2, ZCCHC4, CCDC149, GNPDA2, GUF1, DHX15, SOD3, GUF1 and GNPDA2 as candidate genes. These genes are enriched in signaling pathways such as progesterone-mediated oocyte maturation, parathyroid hormone synthesis, secretion and action, and selenocompound metabolism. ANAPC4, DHX15, GUF1, SEPSECS, and SOD3 were also differently expressed genes (DEGs) in blood RNA-seq. Integrating the GWAS, colocalization, differential expression analysis, and existing literature, ANAPC4 and rs136363104 were highlighted as the putative causal gene and variant for puberty, respectively. In conclusion, our study identified three candidate regions and 12 candidate genes, confirming ANAPC4 as a putative causal gene influencing AFS. This research enhances the understanding of the genetic basis of AFS in Chinese Holstein heifers. The identified key genomic regions, candidate genes, and variants have the potential to improve reproductive efficiency and reduce economic costs in heifer rearing. Interpretative summaryThe economic burden of heifer-raising and reproductive inefficiency in dairy cattle is evident. The most important strategy for optimizing heifer-raising is selecting heifers with earlier puberty (sexual maturity) to shorten age at first service (AFS). In this study, we have identified key genomic regions and candidate genes on AFS in Chinese Holstein heifers by genome-wide association study, and validated them by colocalization and RNA-seq analysis. The findings have the potential to enhance the understanding of the genetic basis of reproductive traits, leading to more targeted breeding strategies for improved reproductive efficiency.

genetics↗