Biology subjects

Yang, J.

Publications and source records attributed to Yang, J..

At least 973 records · Page 54Linked to original sources

bioRxiv2019

Can computers understand word meanings like the human brain does? Comparable semantic representation in neural and computer systems

Semantic representation has been studied independently in neuroscience and computer science. A deep understanding of human neural computations and the revolution to strong artificial intelligence appeal for a joint force in the language domain. We investigated comparable representational formats of lexical semantics between these two complex systems with fine temporal resolution neural recordings. We found semantic representations generated from computational models significantly correlated with EEG responses at an early stage of a typical semantic processing time window in a two-word semantic priming paradigm. Moreover, three representative computational models differentially predicted EEG responses along the dynamics of word processing. Our study provided a finer-grained understanding of the neural dynamics underlying semantic processing and developed an objective biomarker for assessing human-like computation in computational models. Our novel framework trailblazed a promising way to bridge across disciplines in the investigation of higher-order cognitive functions in human and artificial intelligence.

neuroscience↗

bioRxiv2019

A screening platform based on epitope editing for drug discovery

The interaction between an antibody and its epitope has been daily utilized in various biological studies; however it has been rarely explored whether small molecules can alter the interaction. We discovered that small molecules could alter/edit surface properties of amyloid beta (A{beta}) epitopes, and consequently inhibit or enhance corresponding antibody recognition. Remarkably, this editing effect could generate functional changes including protein aggregation behaviors, cell cytokine secreting and in vivo microglia activation. According to this discovery, we proposed a screen platform based on epitope editing for drug discovery (SPEED). With a small library of compounds, we validated that SPEED could be used to seek new leads for A{beta} species. We also demonstrated that this platform could potentially be extended to other targets including tau protein and PD-L1 protein. The SPEED is a simple, fast and label-free screening method. We believe that the SPEED strategy could be universally applicable for seeking and validating drug candidates and imaging ligands.

biochemistry↗

bioRxiv2019

Existence and functions of hypothalamic kisspeptin neuropeptide signaling system in a non-chordate deuterostome species

The kisspeptin (Kp) system is a central modulator of the hypothalamic-pituitary-gonadal axis in vertebrates. Its existence outside the vertebrate lineage remains largely unknown. Here we report the identification and characterization of Kp system in the sea cucumber Apostichopus japonicus. The gene encoding the Kp precursor, generates two mature neuropeptides, AjKiss1a and AjKiss1b. The Kp receptors, AjKissR1 and AjKissR2, are strongly activated by synthetic A. japonicus and vertebrate Kps, triggering a rapid intracellular mobilization of Ca2+, followed by receptor internalization. AjKissR1 and AjKissR2 share similar intracellular signaling pathways via Gq/PLC/PKC/MAPK cascade, when activated by C-terminal decapeptide (AjKiss1b-10). The A. japonicus Kp system functions in mutiple tissues which are closely related to reproduction and metabolism. Overall, our findings uncover for the first time, to our knowledge, the existence and function of the Kp system in a non-chordate species and provide new evidence to support the ancient origin of the hypothalamic neurosecretory system.

evolutionary biology↗

bioRxiv2019

Improved protein structure prediction using predicted inter-residue orientations

The prediction of inter-residue contacts and distances from co-evolutionary data using deep learning has considerably advanced protein structure prediction. Here we build on these advances by developing a deep residual network for predicting inter-residue orientations in addition to distances, and a Rosetta constrained energy minimization protocol for rapidly and accurately generating structure models guided by these restraints. In benchmark tests on CASP13 and CAMEO derived sets, the method outperforms all previously described structure prediction methods. Although trained entirely on native proteins, the network consistently assigns higher probability to de novo designed proteins, identifying the key fold determining residues and providing an independent quantitative measure of the "ideality" of a protein structure. The method promises to be useful for a broad range of protein structure prediction and design problems.

bioinformatics↗

bioRxiv2019

AutoGenome: An AutoML Tool for Genomic Research

Deep learning have made great successes in traditional fields like computer vision (CV), natural language processing (NLP) and speech processing. Those achievements greatly inspire researchers in genomic study and make deep learning in genomics a very hot topic. Convolutional neural network (CNN) and recurrent neural network (RNN) are frequently used for genomic sequence prediction problems; multiple layer perception (MLP) and auto-encoders (AE) are frequently used for genomic profiling data like RNA expression data and gene mutation data. Here, we introduce a new neural network architecture, named residual fully-connected neural network (RFCN) and demonstrate its advantage for modeling genomic profiling data. We further incorporate AutoML algorithms and implement AutoGenome, an end-to-end automated genomic deep learning framework. By utilizing the proposed RFCN architectures, automatic hyper-parameter search and neural architecture search algorithms, AutoGenome can train high-performance deep learning models for various kinds of genomic profiling data automatically. To make researchers better understand the trained models, AutoGenome can assess the feature importance and export the most important features for supervised learning tasks, and the representative latent vectors for unsupervised learning tasks. We envision AutoGenome to become a popular tool in genomic studies.

genomics↗

bioRxiv2019

Quantifying genetic heterogeneity between continental populations for human height and body mass index

Genome-wide association studies (GWAS) in samples of European ancestry have identified thousands of genetic variants associated with complex traits in humans. However, it remains largely unclear whether these associations can be used in non-European populations. Here, we seek to quantify the proportion of genetic variation for a complex trait shared between continental populations. We estimated the between-population correlation of genetic effects at all SNPs (rg) or genome-wide significant SNPs (rg(GWS)) for height and body mass index (BMI) in samples of European (EUR; n = 49,839) and African (AFR; n = 17,426) ancestry. The [Formula] between EUR and AFR was 0.75 (s. e. = 0.035) for height and 0.68 (s. e. = 0.062) for BMI, and the corresponding [Formula] was 0.82 (s. e. = 0.030) for height and 0.87 (s. e. = 0.064) for BMI, suggesting that a large proportion of GWAS findings discovered in Europeans are likely applicable to non-Europeans for height and BMI. There was no evidence that [Formula] differs in SNP groups with different levels of between-population difference in allele frequency or linkage disequilibrium, which, however, can be due to the lack of power.

genetics↗

bioRxiv2019

Simultaneous analysis of open chromatin, promoter interactions and gene expression in stimulated T cells implicates causal genes for rheumatoid arthritis

Genome-wide association studies have identified genetic variation contributing to complex disease risk. However, assigning causal genes and mechanisms has been more challenging because disease-associated variants are often found in distal regulatory regions with cell-type specific behaviours. Here, we collect ATAC-seq, Hi-C, Capture Hi-C and nuclear RNA-seq data in stimulated CD4+ T-cells over 24 hours, to identify functional enhancers regulating gene expression. We characterise changes in DNA interaction and activity dynamics that correlate with changes gene expression, and find that the strongest correlations are observed within 200 kb of promoters. Using rheumatoid arthritis as an example of T-cell mediated disease, we demonstrate interactions of expression quantitative trait loci with target genes, and confirm assigned genes or show complex interactions for 20% of disease associated loci, including FOXO1, which we confirm using CRISPR/Cas9.

genomics↗

bioRxiv2019

RFEX: Simple Random Forest Model and Sample Explainer for non-Machine Learning experts

Machine Learning (ML) is becoming an increasingly critical technology in many areas. However, its complexity and its frequent non-transparency create significant challenges, especially in the biomedical and health areas. One of the critical components in addressing the above challenges is the explainability or transparency of ML systems, which refers to the model (related to the whole data) and sample explainability (related to specific samples). Our research focuses on both model and sample explainability of Random Forest (RF) classifiers. Our RF explainer, RFEX, is designed from the ground up with non-ML experts in mind, and with simplicity and familiarity, e.g. providing a one-page tabular output and measures familiar to most users. In this paper we present significant improvement in RFEX Model explainer compared to the version published previously, a new RFEX Sample explainer that provides explanation of how the RF classifies a particular data sample and is designed to directly relate to RFEX Model explainer, and a RFEX Model and Sample explainer case study from our collaboration with the J. Craig Venter Institute (JCVI). We show that our approach offers a simple yet powerful means of explaining RF classification at the model and sample levels, and in some cases even points to areas of new investigation. RFEX is easy to implement using available RF tools and its tabular format offers easy-to-understand representations for non-experts, enabling them to better leverage the RF technology.

bioinformatics↗

bioRxiv2019

How do we segment text? Two-stage chunking operation in reading

Chunking in language comprehension is a process that segments continuous linguistic input into smaller chunks that are in readers mental lexicon. Effective chunking during reading facilitates disambiguation and enhances efficiency for comprehension. However, the mechanisms of chunking remain elusive, especially in reading given that information arrives simultaneously yet the written systems may not have explicit cues for labeling boundaries such as Chinese. What are the mechanisms of chunking operation that mediates the reading of the text that normally contains hierarchical information? We investigated this question by manipulating the lexical status of the chunks at distinct levels of grain-size in four-character Chinese strings, including the two-character local chunk and four-character global chunk. Participants were asked to make lexical decision on these strings in a behavioral experiment, followed by a passive reading task when their electroencephalography (EEG) were recorded. The behavioral results showed that the lexical decision time of lexicalized two-character local chunks was influenced by the lexical status of four-character global chunk, but not vice versa, which indicated that the processing of global chunks possessed priority over the local chunks. The EEG results revealed that familiar lexical chunks were detected simultaneously at both levels and further processed in a different temporal order -- the onset of lexical access for the global chunks was earlier than that of local chunks. These consistent behavioral and EEG results suggest that chunking in reading occurs at multiple levels via a two-stage operation -- simultaneous detection and global-first recognition.\n\nSignificance StatementThe learners of a new language often read word by word. But why can proficient readers read multiple words at a time? The current study investigates how we efficiently segment a complicate text into smaller pieces and how we process these pieces. Participants read Chinese strings with different structures while their key-press responses and brain EEG signals were recorded. We found that texts were quickly (about 100 ms from their occurrences) segmented to varied sizes of pieces, and larger pieces were then processed earlier than small pieces. Our results suggest that readers can use existing knowledge to efficiently segment and process written information.

animal behavior and cognition↗

bioRxiv2019

Loss of Competence in early development is mediated by loss of chromatin accessibility

As development proceeds, inductive cues are interpreted by competent tissues in a spatially and temporally restricted manner. While key inductive signaling pathways within competent cells are well-described at a molecular level, the mechanisms by which tissues lose responsiveness to inductive signals are not well understood. Localized activation of Wnt signaling before zygotic gene activation in Xenopus laevis leads to dorsal development, but competence to induce dorsal genes in response to Wnts is lost by the late blastula stage. We hypothesize that loss of competence is mediated by changes in histone modifications leading to a loss of chromatin accessibility at the promoters of Wnt target genes. We use ATAC-seq to evaluate genome-wide changes in chromatin accessibility across several developmental stages. Based on overlap with p300 binding, we identify thousands of putative cis-regulatory elements at the gastrula stage, including sites that lose accessibility by the end of gastrulation and are enriched for pluripotency factor binding motifs. Dorsal Wnt target gene promoters are not accessible after the loss of competence in the early gastrula while genes involved in mesoderm and neural crest development maintain accessibility at their promoters. Inhibition of histone deacetylases increases acetylation at the promoters of dorsal Wnt target genes and extends competence for dorsal gene induction by Wnt signaling. Histone deacetylase inhibition, however, is not sufficient to extend competence for mesoderm or neural crest induction. These data suggest that chromatin state regulates the loss of competence to inductive signals in a context-dependent manner.

developmental biology↗

bioRxiv2019

Placental Editome Profiling Reveals Widespread RNA-Editing Dysregulation in Preeclampsia

Dysregulated RNA editing is well documented in several diseases such as cancer. The extent to which RNA editing might be involved in diseases originated in the placenta remains unknown, because RNA editing has rarely been studied in the placenta. Here, we have systematically profiled RNA editome on the placentae from 9 patients with early-onset severe preeclampsia (EOSPE) and 32 normal subjects, and a widespread RNA editing dysregulation in EOSPE has been identified. The mis-edited gene set is enriched with known preeclampsia-associated genes and differentially expressed genes in EOSPE. The \"RNA editing events\" at two microRNA binding sites in 3-UTR of the LEP mRNA were generated, which could lead to increased expression level of LEP in trophoblast cells. Upregulation of LEP were also observed in the placentae of PE patients. These results suggest that widespread placental RNA editing may be involved in placental development and dysregulation of RNA editing in the placenta may contribute to the pathogenesis of preeclampsia.

genomics↗

bioRxiv2019

Towards the identification of causal genes for age-related macular degeneration

Age-related macular degeneration (AMD) is a leading cause of visual impairment in ageing populations and has no radical treatment or prevention. Although genome-wide association studies (GWAS) have identified many susceptibility loci for AMD, the underlying causal genes remain elusive. Here, we prioritized nine putative causal genes by integrating expression quantitative trait locus (eQTL) data from blood (n = 2,765) with AMD GWAS data (16,144 cases vs. 17,832 controls) and replicated six of them using retina eQTL data (n = 523). Of the six genes, altering expression of cnn2, sarm1 and bloc1s1 led to ocular phenotype, impaired vision and retinal pigment epithelium (RPE) loss in zebrafish. Essential photoreceptor and RPE genes were downregulated in cnn2- and sarm1-knockdown zebrafishes. Through integration of GWAS and eQTL data followed by functional validation, our study reveals potential roles of CNN2, SARM1 and BLOC1S1 in AMD pathogenesis and demonstrates an efficient platform to prioritise causal genes for human complex diseases.

genetics↗

bioRxiv2019

Acid-Sensing Ion Channels Mediate Type III Adenylyl Cyclase-Independent Acid-Sensing of Mouse Olfactory Sensory Neurons

Acids can disturb the ecosystem of wild animals through altering their olfaction and olfaction-related survival behaviors. It is known that the main olfactory epithelia (MOE) of mammals rely on odorant receptors and type III adenylyl cyclase (AC3) to detect general odorants. However, it is unknown how the olfactory system sense protons or acidic odorants. Here we show that the mouse MOE responded to acidic volatile stimuli in the presence and the absence of AC3. Acetic acid-induced electro-olfactogram (EOG) responses in wild type (WT) MOE can be dissected into two components: one dependent on the AC3-mediated cAMP pathway and the other not. MOE of AC3 knockout (KO) mice retained an acid-evoked EOG response but failed to respond to an odor mix. The acid-evoked responses of the AC3 KO could be blocked by diminazene, an inhibitor of acid-sensing ion channels (ASICs), but not by forskolin/IBMX, which desensitize the cAMP pathway. AC3 KO mice lost their sensitivity to detect pungent odorants but maintained sniffing behavior to acetic acid. Immunofluorescence staining demonstrated that ASIC1 proteins were highly expressed in olfactory sensory neurons (OSNs), mostly enriched in the knobs, dendrites, and somata, but not in olfactory cilia. Moreover, mice exhibited reduced preference to attractive objects when placed in an environment with acidic volatiles. Together, we conclude that the mouse olfactory system has a non-conventional, ASICs-mediated mechanism for acid-sensing. Acid stimulation of ASICs may unselectively depolarize different OSNs and interfere with the anatomical logic for odor perception.

neuroscience↗

bioRxiv2019

Epitope containing short peptides capture distinct IgG serodynamics that enable DIVA for live-attenuated vaccines

Differentiating infected from vaccinated animals (DIVA) strategies have been central enabling techniques in several successful viral disease elimination programs. However, owing to their long and uncertain development process, no DIVA-compatible vaccines are available for many important diseases. We report herein a new DIVA strategy based on hybrid protein-peptide microarrays which can theoretically work with any vaccine. Leading from our findings from Peste des petits ruminants (PPR), we found 4 epitope containing short peptides (ECSPs) which have distinct IgG serodynamics: anti-ECSP IgGs only exist for 10-60 days post vaccination (dpv), while anti-protein IgGs remained at high levels for >1000 dpv. These data enabled design of a DIVA diagnostic microarray containing 4 ECSPs and 3 proteins, which unlike cELISA and VNT, enables ongoing monitoring of serological differences between vaccinated individuals and individuals exposed to the pathogen. For 50 samples after 60 dpv, 20 animals were detected with positive anti-ECSP IgGs, indicating recent infections in vaccinated goat/sheep herds. These DIVA diagnostic microarrays will almost certainly facilitate eradication programs for (re-)emerging pathogens and zoonoses.

immunology↗

bioRxiv2019

The Influence of Blood Collection Tubes in Biomarkers Screening by Mass Spectrometry

BackgroundMass spectrometry (MS) is one of the rapidly developing bio-analytical techniques in recent years and have found many biomarkers of variety of disease. Whereas pre-analytical process is one of most crucial procedure which would significantly influence the results of biomarkers screening. In the current study, we conducted a pilot analysis of serum to determine the effects of blood collection tubes in biomarkers screening.\n\nMethodsMagnetic bead separation and matrix-assisted laser desorption ionization time-of-flight mass spectrometry were used for qualitative analysis of healthy control and serum cancer patients. A total of 24 serum samples were analyzed in this study, of which were collected from patients or healthy control using non-additive tubes or coagulant activator tubes respectively. ClinProTools were used to compare the difference among the different groups.\n\nResultsThe results demonstrated that no matter for patients or normal people, the serum protein profile changed significantly when using coagulant tubes. We also found that the effect of coagulant on serum protein of patients was smaller than that of control group. There were significant differences among 27 peaks which were obtained in the control group and the control coagulant group. However, between patient group and patient coagulant group, only 1 differential peak were obtained. Coagulant changed the protein expression difference in the original serum, and the difference expanded, narrowed even reversed, most of which are small polypeptides (Mass<3000 Da), which significantly changed the results of biomarkers screening. The results showed that 19 potential biomarkers could be found with non-additive tubes and 16 potential biomarkers could be found with coagulate activator tubes, among which only 6 were the same.\n\nConclusionsThe choice of blood collection tube significantly influence the results of biomarkers screening by MS.

cancer biology↗

bioRxiv2019

Computational modeling reveals a hydrophobic force-induced pore-forming mechanism for cholesterol-dependent cytolysins

During the pore-forming process, cholesterol-dependent cytolysins (CDCs) bind to cholesterol-rich membranes and subsequently undergo a series of conformational changes, predominantly involving in the collapse of the protein and the transformation from helices to {beta}-hairpins to form a large hydrophobic pore. In the current study, we reconstructed a structural model for both the prepore and pore-forming complexes of PFO based on the cryo-EM data of pneumolysin and performed molecular dynamics simulations and free energy calculations to study the conformational changes in the PFO prepore-to-pore conversion. Our simulations indicate that D2 cannot collapse spontaneously due to the hydrogen bonding and pi-pi interactions between domains D2 and D3, which are partially weakened by binding to cell membranes and oligomerization. The free energy landscape for the prepore-to-pore conversion reveals that an additional force is required for the collapse of D2 to overcome an energy barrier of [~] 24 kcal/mol. Based on these computational results, we proposed a hydrophobic force-induced pore-forming mechanism to explain the pore-forming process of CDCs. In this mechanism, the hydrophobic interactions between the TMHs and membranes are essential for the prepore-to-pore conversion. The hydrophobic force generated by the TMHs-membrane interactions drives the conformational changes in domains D2 and D3. These findings well explain how the conformational changes within two distant domains synergistically occur, and fits well for the previous biophysical and biochemical data.

biophysics↗

bioRxiv2019

KIT ligand protects against both light-induced and genetic photoreceptor degeneration

Photoreceptor cell degeneration is a major cause of blindness and a considerable health burden during aging but effective therapeutic or preventive strategies have not so far become commercially available. Here we show in mouse models that signaling through the tyrosine kinase receptor KIT protects photoreceptor cells against both light-induced and inherited retinal degeneration. Upon light damage, photoreceptor cells upregulate Kit ligand (KITL) and activate KIT signaling, which in turn induces nuclear accumulation of the transcription factor NRF2 and stimulates the expression of the antioxidant gene Hmox1. Conversely, a viable Kit mutation promotes light-induced photoreceptor damage, which is reversed by experimental expression of Hmox1. Furthermore, overexpression of KITL from a viral AAV8 vector prevents photoreceptor cell death and partially restores retinal function after light damage or in genetic models of human retinitis pigmentosa. Hence, application of KITL may provide a novel therapeutic avenue for prevention or treatment of retinal degenerative diseases.

neuroscience↗

bioRxiv2019

Bayesian analysis of GWAS summary data reveals differential signatures of natural selection across human complex traits and functional genomic categories

Understanding how natural selection has shaped the genetic architecture of complex traits and diseases is of importance in medical and evolutionary genetics. Bayesian methods have been developed using individual-level data to estimate multiple features of genetic architecture, including signatures of natural selection. Here, we present an enhanced method (SBayesS) that only requires GWAS summary statistics and incorporates functional genomic annotations. We analysed GWAS data with large sample sizes for 155 complex traits and detected pervasive signatures of negative selection with diverse estimates of SNP-based heritability and polygenicity. Projecting these estimates onto a map of genetic architecture obtained from evolutionary simulations revealed relatively strong natural selection on genetic variants associated with cardiorespiratory and cognitive traits and relatively small number of mutational targets for diseases. Averaging across traits, the joint distribution of SNP effect size and MAF varied across functional genomic regions (likely to be a consequence of natural selection), with enrichment in both the number of associated variants and the magnitude of effect sizes in regions such as transcriptional start sites, coding regions and 5- and 3-UTRs.

genetics↗