Biology subjects

Gao, Z.

Publications and source records attributed to Gao, Z..

At least 127 records · Page 7Linked to original sources

bioRxiv2024

INSTINCT: Multi-sample integration of spatial chromatin accessibility sequencing data via stochastic domain translation

Recent advances in spatial epigenomic techniques have given rise to spatial assay for transposase-accessible chromatin using sequencing (spATAC-seq) data, enabling the characterization of epigenomic heterogeneity and spatial information simultaneously. Integrative analysis of multiple spATAC-seq samples, for which no method has been developed, allows for effective identification and elimination of unwanted non-biological factors within the data, enabling comprehensive exploration of tissue structures and providing a holistic epigenomic landscape, thereby facilitating the discovery of biological implications and the study of regulatory processes. In this article, we present INSTINCT, a method for multi-sample INtegration of Spatial chromaTIN accessibility sequencing data via stochastiC domain Translation. INSTINCT can efficiently handle the high dimensionality of spATAC-seq data and eliminate the complex noise and batch effects of samples from different conditions through a stochastic domain translation procedure. We demonstrate the superiority and robustness of INSTINCT in integrating spATAC-seq data across multiple simulated scenarios and real datasets. Additionally, we highlight the advantages of INSTINCT in spatial domain identification, visualization, spot-type annotation, and various downstream analyses, including motif enrichment analysis, expression enrichment analysis, and partitioned heritability analysis.

bioinformatics↗

bioRxiv2024

Reduced temporal and spatial stability of neural activity patterns predict cognitive control deficits in children with ADHD

This study explores the neural underpinnings of cognitive control deficits in ADHD, focusing on overlooked aspects of trial-level variability of neural coding. We employed a novel computational approach to neural decoding on a single-trial basis alongside a cued stop-signal task which allowed us to distinctly probe both proactive and reactive cognitive control. Typically developing (TD) children exhibited stable neural response patterns for efficient proactive and reactive dual control mechanisms. However, neural coding was compromised in children with ADHD. Children with ADHD showed increased temporal variability and diminished spatial stability in neural responses in salience and frontal-parietal network regions, indicating disrupted neural coding during both proactive and reactive control. Moreover, this variability correlated with fluctuating task performance and with more severe symptoms of ADHD. These findings underscore the significance of modeling single-trial variability and representational similarity in understanding distinct components of cognitive control in ADHD, highlighting new perspectives on neurocognitive dysfunction in psychiatric disorders.

neuroscience↗

bioRxiv2024

Pi-PrimeNovo: An Accurate and Efficient Non-Autoregressive Deep Learning Model for De Novo Peptide Sequencing

Peptide sequencing via tandem mass spectrometry (MS/MS) is fundamental in proteomics data analysis, playing a pivotal role in unraveling the complex world of proteins within biological systems. In contrast to conventional database searching methods, deep learning models excel in de novo sequencing peptides absent from existing databases, thereby facilitating the identification and analysis of novel peptide sequences. Current deep learning models for peptide sequencing predominantly use an autoregressive generation approach, where early errors can cascade, largely affecting overall sequence accuracy. And the usage of sequential decoding algorithms such as beam search suffers from the low inference speed. To address this, we introduce{pi} -PrimeNovo, a non-autoregressive Transformer-based deep learning model designed to perform accurate and efficient de novo peptide sequencing. With the proposed novel architecture,{pi} -PrimeNovo achieves significantly higher accuracy and up to 69x faster sequencing compared to the state-of-the-art methods. This remarkable speed makes it highly suitable for computation-extensive peptide sequencing tasks such as metaproteomic research, where{pi} -PrimeNovo efficiently identifies the microbial species-specific peptides. Moreover,{pi} -PrimeNovo has been demonstrated to have a powerful capability in accurately mining phosphopeptides in a non-enriched phosphoproteomic dataset, showing an alternative solution to detect low-abundance post-translational modifications (PTMs). We suggest that this work not only advances the development of peptide sequencing techniques but also introduces a transformative computational model with wide-range implications for biological research.

bioinformatics↗

bioRxiv2024

InstructPLM: Aligning Protein Language Models to Follow Protein Structure Instructions

Large language models are renowned for their efficacy in capturing intricate patterns, including co-evolutionary relationships, and underlying protein languages. However, current methodologies often fall short in illustrating the emergence of genomic insertions, duplications, and insertion/deletions (indels), which account for approximately 14% of human pathogenic mutations. Given that structure dictates function, mutated proteins with similar structures are more likely to persist throughout biological evolution. Motivated by this, we leverage crossmodality alignment and instruct fine-tuning techniques inspired by large language models to align a generative protein language model with protein structure instructions. Specifically, we present a method for generating variable-length and diverse proteins to explore and simulate the complex evolution of life, thereby expanding the repertoire of options for protein engineering. Our proposed protein LM-based approach, InstructPLM, demonstrates significant performance enhancements both in silico and in vitro. On native protein backbones, it achieves a perplexity of 2.68 and a sequence recovery rate of 57.51, surpassing Protein-MPNN by 39.2% and 25.1%, respectively. Furthermore, we validate the efficacy of our model by redesigning PETase and L-MDH. For PETase, all fifteen designed variable-length PETase exhibit depolymerization activity, with eleven surpassing the activity levels of the wild type. Regarding L-MDH, an enzyme lacking an experimentally determined structure, InstructPLM is able to design functional enzymes with an AF2-predicted structure. Code and model weights of InstructPLM are publicly available*.

bioinformatics↗

bioRxiv2024

GABAAR-PPT1 palmitoylation homeostasis controls synaptic transmission and circuitry oscillation in CLN1 disease

CLN1 disease, also called infantile neuronal ceroid lipofuscinosis, is a fatal neurodegenerative disease caused by mutations in the CLN1 gene encoding palmitoyl protein thioesterase 1 (PPT1). To identify depalmitoylation substrate of PPT1 is crucial to understand CLN1 disease. In this study, we found that PPT1 depalmitoylates GABAAR 1 subunit at Cystein-260, while binding to Cystein-165 and -179. Mutations of PPT1 or its GABAAR 1 subunit binding site result in enhanced inhibitory synaptic transmission, strengthened and oscillation but disrupted phase coupling in CA1 region and impaired learning and memory in 1- to 2-months-old PPT1-deficient and Gabra1em1 mice. Our study highlights the critical role of PPT1 in maintaining GABAAR palmitoylation homeostasis and reveals a previously unknown molecular pathway in PPT1 induced diseases.

neuroscience↗

bioRxiv2024

collectNET: a web server for integrated inference of cell-cell communication network

SummaryCell-cell communication through ligand-receptor pairs forms the cornerstone for complex functionalities in multicellular organisms. Deciphering such intercellular signaling can contribute to un-raveling disease mechanisms and enables targeted therapy. Nonetheless, notable biases and inconsistencies are evident among the inferential outcomes generated by current methods for inferring cell-cell communication network. To fill this gap, we developed collectNET (http://health.tsing-hua.edu.cn/collectnet) as the first web server for efficiently inferring the cell-cell communication network, with efficient calculation, hierarchical browsing, comprehensive statistics, advanced searching, and intuitive visualization. collectNET provides a reliable online inference service with prior knowledge of three public ligand-receptor databases and systematic integration of three mainstream inference methods. Additionally, collectNET has assembled a human cell-cell communication atlas, including 126,785 significant communication pairs based on 343,023 single cells. We anticipate that collectNET will benefit researchers in gaining a more holistic understanding of cell development and differentiation mechanisms. Availability and ImplementationcollectNET is freely available at http://health.tsinghua.edu.cn/collectnet. Contactruijiang@tsinghua.edu.cn Supplementary informationSupplementary data are available at Bioinformatics online.

genomics↗

bioRxiv2024

In-depth investigation of the mechanisms of high and low residual feed intake regulating hens during the late laying period via liver and gut microbiota.

Residual feed intake (RFI) is a more accurate indicator of feed efficiency than the feed conversion ratio (FCR) and is widely used to measure the efficiency of livestock and poultry feed utilization. Typically, Low RFI (LRFI) implies higher feed conversion efficiency, while high RFI (HRFI) indicates lower feed conversion efficiency. This study systematically explored the differences between high and low RFI and the function of the liver and cecum microbes of hens during the late laying period by multiple-omics techniques and further explored the interaction among microorganisms, the function of tissues and organs, and body metabolism. The results showed that the length and mass of the digestive organs in the LRFI group were higher than those in the HRFI group as well as the chest width. Additionally, the key genes and metabolites regulating RFI in hens during the late laying phase were found to be ADCY2, ADCY8, CCKAR, ACSS2, FABP1, FABP4, and LysoPI (18:2(9Z,12Z)/0:0) in the liver. The levels of AST, HDL-C and ACTH in the serum were considered candidate markers influencing RFI. By conducting a microbiome-metabolome association analysis, we have identified the dominant and beneficial microbial community in the gut of LRFI groups, such as Oscillospirales, Ruminococcaceae, and Butyricicoccaceae, which offers a theoretical basis for understanding how the gut microbiota regulates RFI. These results will provide a scientific basis for the molecular mechanism of RFI phenotypic variation in late laying hens.

microbiology↗

bioRxiv2024

Biosurfer for systematic tracking of regulatory mechanisms leading to protein isoform diversity

Long-read RNA sequencing has shed light on transcriptomic complexity, but questions remain about the functionality of downstream protein products. We introduce Biosurfer, a computational approach for comparing protein isoforms, while systematically tracking the transcriptional, splicing, and translational variations that underlie differences in the sequences of the protein products. Using Biosurfer, we analyzed the differences in 32,799 pairs of GENCODE annotated protein isoforms, finding a majority (70%) of variable N-termini are due to the alternative transcription start sites, while only 9% arise from 5 UTR alternative splicing. Biosurfers detailed tracking of nucleotide-to-residue relationships helped reveal an uncommonly tracked source of single amino acid residue changes arising from the codon splits at junctions. For 17% of internal sequence changes, such split codon patterns lead to single residue differences, termed "ragged codons". Of variable C-termini, 72% involve splice- or intron retention-induced reading frameshifts. We found an unusual pattern of reading frame changes, in which the first frameshift is closely followed by a distinct second frameshift that restores the original frame, which we term a "snapback" frameshift. We analyzed long read RNA-seq-predicted proteome of a human cell line and found similar trends as compared to our GENCODE analysis, with the exception of a higher proportion of isoforms predicted to undergo nonsense-mediated decay. Biosurfers comprehensive characterization of long-read RNA-seq datasets should accelerate insights of the functional role of protein isoforms, providing mechanistic explanation of the origins of the proteomic diversity driven by the alternative splicing. Biosurfer is available as a Python package at https://github.com/sheynkman-lab/biosurfer.

genomics↗

bioRxiv2024

Cas1 mediates the interference stage in a phage-encoded CRISPR-Cas system

CRISPR-Cas systems are prokaryotic adaptive immune systems against invading phages and other mobile genetic elements, which function in three stages: adaptation, expression and interference. Interestingly, phages were also found to encode CRISPR-Cas systems to antagonize their hosts and establish infection. The Vibrio cholerae-infecting ICP1 phage was the first discovered phage with a functional CRISPR-Cas system against an antiphage region in the host genome. Nevertheless, this system lacks a domain essential for recruitment of helicase-nuclease Cas2/3 during target DNA cleavage, and how this system accomplishes the interference stage remains unknown. Here, surprisingly, we found that Cas1, a highly conserved component known to exclusively work in the adaptation stage, also mediates the interference stage through connecting Cas2/3 to the DNA bound-Cascade (Csy) complex of the ICP1 CRISPR-Cas system. A series of structures of Csy, Csy-dsDNA, Cas1-Cas2/3 and Csy-dsDNA-Cas1-Cas2/3 complexes collectively reveal the whole process of Cas1-mediated target DNA cleavage by the ICP1 CRISPR-Cas system. Together, these data support an unprecedented model in which Cas1 mediates the interference stage in a phage-encoded CRISPR-Cas system and also shed light on a unique model of primed adaptation.

microbiology↗

bioRxiv2024

Graspot: A graph attention network for spatial transcriptomics data integration with optimal transport

Spatial transcriptomics (ST) technologies enable the measurement of mRNA expression while simultaneously capturing spot locations. By integrating ST data, the 3D structure of a tissue can be reconstructed, yielding a comprehensive understanding of the tissues intricacies. Nevertheless, a computational challenge persists: how to remove batch effects while preserving genuine biological structure variations across ST data. To address this, we introduce Graspot, a graph attention network designed for spatial transcriptomics data integration with unbalanced optimal transport. Graspot adeptly harnesses both gene expression and spatial information to align common structures across multiple ST datasets. It embeds multiple ST datasets into a unified latent space, facilitating the partial alignment of spots from different slices. Demonstrating superior performance compared to existing methods on four real spatial transcriptomics datasets, Graspot excels in ST data integration, including tasks that require partial alignment. In particular, Graspot unveils subtle tumor microenvironment structures of breast cancer, and accurately aligns the spatio-temporal transcriptomics data to reconstruct human heart developmental processes. The code for Graspot is available at https://github.com/zhan009/Graspot.

bioinformatics↗

bioRxiv2024

Self-amplifying ROS-sensitive SN38 Dimeric Prodrug nanoparticles for Combined Chemotherapy and Ferroptosis in Cancer Treatment

Chemotherapy drugs are often limited by their own clinical shortcomings and serious adverse consequences. To solve these problems, we developed a self-amplifying reactive oxygen species (ROS)-sensitive dimeric prodrug nanoparticles, namely SN38-CA@FC NPs for tumor treatment. A ROS-sensitive 7-ethyl-10-hydroxycamptothecin (SN38) prodrug (SN38-CA) was synthesized by a thioacetal linker between SN38 and the ROS generator cinnamaldehyde (CA). The subsequent release of SN38 inflicts DNA damage, exerting chemotherapeutic effects, while the liberated CA intensifies ferroptosis through Fenton reaction-mediated disruption of the redox balance. This dual-action strategy not only leverages chemotherapy but also induces ferroptosis, establishing a synergistic therapeutic paradigm. The system is uniquely characterized by a positive feedback loop where ROS instigates the release of SN38/CA, which in turn promotes further ROS production. In experimental evaluations, this combination therapy exhibited potent antitumor activity against both A549 and LLC cancer cell lines, as well as in xenograft LLC-bearing C57BL/6 mouse models. Collectively, our findings introduce a transformative Nano-Drug Delivery System (NDDS) that holds significant promise for advancing cancer chemotherapy and ferroptosis-based therapies.

cancer biology↗

bioRxiv2024

Evaluation of the effects of ripening-related genes on fruit metabolites and the associated regulatory mechanisms in tomato

Fruit ripening, which is a complex process involving dynamic changes to metabolites, is controlled by multiple factors, including transcription factors (TFs). Several TFs are reportedly essential regulators of tomato fruit ripening. To evaluate the effects of specific TFs on metabolite accumulation during fruit ripening, CRISPR/Cas9-mediated mutagenesis was combined with metabolome and transcriptome analyses to explore regulatory mechanisms. Specifically, we generated various genetically engineered tomato lines that differed regarding metabolite contents and fruit colors. The metabolite and transcript profiles indicated that the selected TFs have distinct functions that control fruit metabolite contents, especially carotenoids and sugars. Our findings may provide new insights into the regulatory mechanisms governing tomato fruit ripening. Moreover, a mutation to ELONGATED HYPOCOTYL5 (HY5) increased the tomato fruit fructose and glucose contents by approximately 20% (relative to the wild-type levels). Our in vitro assay showed that HY5 can bind directly to the G-box cis-element in the SWEET12c promoter to activate expression, thereby modulating sugar transport. Our findings have clarified the mechanism regulating fruit metabolic networks, while also providing the theoretical basis for breeding horticultural crops that produce fruit with diverse flavors and colors.g

plant biology↗

bioRxiv2023

Discrimination of Isomeric Glycans via High-sensitive Engineered Nanopore

Glycan isomers are the basis for forming the spatial structure of glycans, which poses great challenges to glycan structural analysis and hinders the study of the relationship between structure and function. An economical and convenient analytical method with high resolution is still in demand. Here, we designed an engineered -hemolysin nanopore with double-mutants (-HL M113R/T115A) to achieve sensing of glycans with different glycosidic linkages. By extracting three parameters to depict a 3D scatter plot, four glycans with different degrees of polymerization and glycosidic bonds could be distinguished. Through molecular dynamics (MD), we further elucidated the motion trajectories of two different glycosidic bonds in the pore region. In addition, the sensing ability was also applied in the direct identification of glycans in mixture systems. The obtained nanopore data permitted us to achieve a separation accuracy of over 90% independent of the assistance of machine learning algorithms. Designing biological nanopores with high discrimination and sensitivity for glycans with different lineages paves the way for nanopore glycan profiling and potentially sequencing.

bioengineering↗

bioRxiv2023

Visual to default network pathways: A double dissociation between semantic and spatial cognition

Processing pathways between sensory and default mode network (DMN) regions support recognition, navigation, and memory but their organisation is not well understood. We show that functional subdivisions of visual cortex and DMN sit at opposing ends of parallel streams of information processing that support visually-mediated semantic and spatial cognition, providing convergent evidence from univariate and multivariate task responses, intrinsic functional and structural connectivity. Participants learned virtual environments consisting of buildings populated with objects, drawn from either a single semantic category or multiple categories. Later, they made semantic and spatial context decisions about these objects and buildings during functional magnetic resonance imaging. A lateral ventral occipital to frontotemporal DMN pathway was primarily engaged by semantic judgements, while a medial visual to medial temporal DMN pathway supported spatial context judgements. These pathways had distinctive locations in functional connectivity space: the semantic pathway was both further from unimodal systems and more balanced between visual and auditory-motor regions compared with the spatial pathway. When semantic and spatial context information could be integrated (in buildings containing objects from a single category), regions at the intersection of these pathways responded, suggesting that parallel processing streams interact at multiple levels of the cortical hierarchy to produce coherent memory-guided cognition.

neuroscience↗

bioRxiv2023

Neuronal dynamics of cerebellum and medial prefrontal cortex in adaptive motor timing

Precise temporal control of sensorimotor coordination and adaptation is a fundamental basis of animal behavior. How different brain regions are involved in regulating the flexible temporal adaptation remains elusive. Here we investigated the neuronal dynamics of cerebellar interposed nucleus (IpN) and medial prefrontal cortex (mPFC) neurons during temporal adaptation between delay eyeblink conditioning (DEC) and trace eyeblink conditioning (TEC). When mice trained for either DEC or TEC and subsequently subjected to a new paradigm, their conditioned responses (CRs) adapted virtually instantaneously. Changes in the activity of the IpN neurons to CR timing were prominent during DEC-to-TEC adaptation, but less so during TEC-to-DEC adaptation. In contrast, mPFC neurons could rapidly alter their modulation patterns during both adaptation paradigms. Accordingly, silencing of mPFC blocked the adaptation of CR timing. These results illuminate how cerebral and cerebellar mechanisms may play differential roles during adaptive control of associative motor timing.

neuroscience↗

bioRxiv2023

Human microbiota is a reservoir of SARS-CoV-2 advantageous mutations

SARS-CoV-2 mutations are rapidly emerging, in particular advantageous mutations in the spike (S) protein, which either increase transmissibility or lead to immune escape, are posing a major challenge to pandemic prevention and treatment. However, how the virus acquires a high number of advantageous mutations in a short time remains a mystery. Here, we show that the human microbiota may contribute to mutations in variants of concern (VOCs). We identified a mutation and adjacent 6 amino acids (aa) in a viral mutation fragment (VMF) and searched for homologous fragments (HFs) in the National Center for Biotechnology Information (NCBI) database. Among the approximate 8000 HFs obtained, 61 mutations in S and other outer membrane proteins were found in bacteria, accounting for 62% of all mutation sources, which is a 12-fold higher than the natural variable proportion. Approximately 70% of these bacterial species belong to the human microbiota, are primarily distributed in the gut or lung and exhibit a composition pattern similar to that of COVID-19 patients. Importantly, SARS- CoV-2 RNA-dependent RNA polymerase (RdRp) replicates corresponding bacterial mRNAs harboring mutations, producing chimeric RNAs. Collectively, SARS-CoV-2 may acquire mutations from the human microbiota, resulting in alterations in the binding sites or antigenic determinants of the original virus. Our study sheds light on the evolving mutational mechanisms of SARS-CoV-2.

microbiology↗

bioRxiv2023

Single phage proteins sequester TIR- and cGAS-generated signaling molecules

Prokaryotic anti-phage immune systems use TIR (toll/interleukin-1 receptor) and cGAS (cyclic GMP-AMP synthase) enzymes to produce 1-3/1-2 glycocyclic ADPR (gcADPR) and cyclid di-/tri-nucleotides (CDNs and CTNs) signaling molecules that limit phage replication, respectively 1-3. However, how phages neutralize these common systems is largely unknown. Here, we show that Thoeris anti-defense proteins Tad1 4 and Tad2 5 both have anti-CBASS activity by simultaneously sequestering CBASS cyclic oligonucleotides. Strikingly, apart from binding Thoeris signals 1-3 and 1-2 gcADPR, Tad1 also binds numerous CBASS CDNs/CTNs with high affinity, inhibiting CBASS systems using these molecules in vivo and in vitro. The hexameric Tad1 has six binding sites for CDNs or gcADPR, which are independent from two high affinity binding sites for CTNs. Tad2 also sequesters various CDNs in addition to gcADPR molecules, inhibiting CBASS systems using these CDNs. However, the binding pockets for CDNs and gcADPR are different in Tad2, whereby a tetramer can bind two CDNs and two gcADPR molecules simultaneously. Taken together, Tad1 and Tad2 are both two-pronged inhibitors that, alongside anti-CBASS protein 2, establish a paradigm of phage proteins that flexibly sequester a remarkable breadth of cyclic nucleotides involved in TIR- and cGAS-based anti-phage immunity.

microbiology↗

bioRxiv2023

Structural determinants of pure antiestrogenicity

Inhibitory activities of antiestrogens on estrogen receptor alpha (ER) range from mixed antagonism/agonism (selective ER modulators; SERMs) to complete antiestrogenicity associated with accelerated ER turnover (selective ER degraders; SERDs). Here, we show using a panel of SERMs, SERDs and a PROTAC that efficient induction of ER SUMOylation better accounts for complete transcriptional repression than increased suppression of coactivator recruitment or accelerated ER degradation. Antiestrogen-induced ER SUMOylation depends on the hydrophobicity of N-terminal residues of ligand binding domain (LBD) helix 12 (H12). L536 mutations, including those occurring in endocrine therapy-resistant breast cancer, abolished SUMOylation with all antiestrogens. Structures of the L536S human ER LBD bound to fulvestrant analogs and molecular dynamics simulations predict dynamic side chain interactions with wild-type ER H12 in the coactivator-binding groove and model the impact of L536 mutations. Together, our results show that induced SUMOylation contributes to the complete suppression of ER activity and provide insights into its structural determinants.

molecular biology↗