Biology subjects

Zheng, G.

Publications and source records attributed to Zheng, G..

At least 55 records · Page 3Linked to original sources

bioRxiv2023

Methods for constructing and evaluating consensus genomic interval sets

The amount of genomic region data continues to increase. Integrating across diverse genomic region sets requires consensus regions, which enable comparing regions across experiments, but also by necessity lose precision in region definitions. We require methods to assess this loss of precision and build optimal consensus region sets. Here, we introduce the concept of flexible intervals and propose 3 novel methods for building consensus region sets, or universes: a coverage cutoff method, a likelihood method, and a Hidden Markov Model. We then propose 3 novel measures for evaluating how well a proposed universe fits a collection of region sets: a base-level overlap score, a region boundary distance score, and a likelihood score. We apply our methods and evaluation approaches to several collections of region sets and show how these methods can be used to evaluate fit of universes and build optimal universes. We describe scenarios where the common approach of merging regions to create consensus leads to undesirable outcomes and provide principled alternatives that provide interoperability of interval data while minimizing loss of resolution. Software is available at https://github.com/databio/geniml.

genomics↗

bioRxiv2023

Fast clustering and cell-type annotation of scATAC data using pre-trained embeddings

MotivationData from the single-cell assay for transposase-accessible chromatin using sequencing (scATAC-seq) is now widely available. One major computational challenge is dealing with high dimensionality and inherent sparsity, which is typically addressed by producing lower-dimensional representations of single cells for downstream clustering tasks. Current approaches produce such individual cell embeddings directly through a one-step learning process. Here, we propose an alternative approach by building embedding models pre-trained on reference data. We argue that this provides a more flexible analysis workflow that also has computational performance advantages through transfer learning. ResultsWe implemented our approach in scEmbed, an unsupervised machine learning framework that learns low-dimensional embeddings of genomic regulatory regions to represent and analyze scATAC-seq data. scEmbed performs well in terms of clustering ability and has the key advantage of learning patterns of region co-occurrence that can be transferred to other, unseen datasets. Moreover, pre-trained models on reference data can be exploited to build fast and accurate cell-type annotation systems without the need for other data modalities. scEmbed is implemented in Python and it is available to download from GitHub. We also make our pre-trained models available on huggingface for public use. AvailabilityscEmbed is open source and available at https://github.com/databio/geniml. Pre-trained models from this work can be obtained on huggingface: https://huggingface.co/databio.

genomics↗

bioRxiv2023

Dietary fibre reverses adverse post-stroke outcomes in mice via short-chain fatty acids and its sensing receptors GPR41, GPR43 and GPR109A

Dietary fibre intake is associated with fewer cases of ischaemic stroke. This is likely via the microbiota-gut-brain axis, where fibre is fermented by the gut microbiota, releasing short-chain fatty acids (SCFAs). However, whether fibre or SCFAs can reverse adverse post-stroke outcomes remains unknown. Here, we demonstrated that a low fibre diet exacerbates post-stroke outcomes in mice. This was reversed by a high fibre diet or direct supplementation with SCFAs (delivered either in the water or a high SCFA-releasing diet) immediately after stroke. These modulated the gut microbiome and improved the gut epithelial barrier integrity, which was associated with fewer activated neutrophils and more neuroblast cells in the brain. We then investigated the SCFA-receptors GPR41/43/109A using a triple knockout mouse model, which exhibited poorer stroke outcomes and recovery. These results show that post-stroke interventions using dietary fibre and/or SCFA supplementation, acting via GPR41/43/109A signalling, may represent new therapeutic strategies for stroke-induced brain injury.

neuroscience↗

bioRxiv2022

SBT-272 improves TDP-43 pathology in the ALS motor cortex by modulating mitochondrial integrity, motility, and function

Mitochondrial defects are one of the common underlying causes of neuronal vulnerability in motor neuron diseases, such as amyotrophic lateral sclerosis (ALS), and TDP-43 pathology is the most common proteinopathy in ALS. Disrupted inner mitochondrial membrane (IMM) reported in the upper motor neurons (UMNs) of ALS patients with TDP-43 pathology is recapitulated in the UMNs of well-characterized mutant hTDP-43 mouse models of ALS. The construct validity, such as common cellular pathology in mice and human, offers a unique opportunity to test treatment strategies that may translate. SBT-272 is a well-tolerated brain-penetrant small molecule that stabilizes cardiolipin, a phospholipid found in IMM, thereby restoring mitochondrial structure and respiratory function. We investigated whether SBT-272 can improve IMM structure and health in UMNs diseased with TDP-43 pathology in our well-characterized UMN reporter line for ALS. We found that SBT-272 significantly improved mitochondrial structural integrity and restored mitochondrial motility and function. This led to improved health of diseased UMNs in vitro. In comparison to edaravone and AMX0035, SBT-272 appeared more effective in restoring health of diseased UMNs. Chronic treatment of SBT-272 for sixty days starting at an early symptomatic stage of the disease in vivo led to a reduction in astrogliosis, microgliosis, and retention of UMN degeneration in the ALS motor cortex. Our results underscore the therapeutic potential of SBT-272, especially within the context of TDP-43 pathology and mitochondrial dysfunction. HighlightsO_LIEarly and progressive upper motor neuron (UMN) degeneration defines ALS pathology C_LIO_LIMitochondrial defects are prominent and common in UMNs with TDP-43 pathology C_LIO_LISBT-272 treatment improves mitochondrial stability, mobility and function C_LIO_LISBT-272 treatment reduces astrogliosis, microgliosis and improves UMN health C_LI

neuroscience↗

bioRxiv2022

Co-targeting BCL-XL and MCL-1 with DT2216 and AZD8055 synergistically inhibits small-cell lung cancer growth without causing on-target toxicities in mice

Small-cell lung cancer (SCLC) is an aggressive malignancy with limited therapeutic options. The dismal prognosis in SCLC is in part associated with an upregulation of BCL-2 family anti-apoptotic proteins, including BCL-XL and MCL-1. Unfortunately, the currently available inhibitors of BCL-2 family anti-apoptotic proteins, except BCL-2 inhibitors, are not clinically relevant because of various on-target toxicities. We, therefore, aimed to develop an effective and safe strategy targeting these anti-apoptotic proteins with DT2216 (our platelet-sparing BCL-XL degrader) and AZD8055 (an mTOR inhibitor) to avoid associated on-target toxicities while synergistically optimizing tumor response. Through BH3 mimetic screening, we identified a subset of SCLC cell lines that is co-dependent on BCL-XL and MCL-1. After screening inhibitors of selected tumorigenic pathways, we found that AZD8055 selectively downregulates MCL-1 in SCLC cells and its combination with DT2216 synergistically killed BCL-XL/MCL-1 co-dependent SCLC cells, but not normal cells. Mechanistically, the combination caused BCL-XL degradation and suppression of MCL1 expression, and thus disrupted MCL-1 interaction with BIM leading to an enhanced apoptotic induction. In vivo, DT2216+AZD8055 combination significantly inhibited the growth of cell line-derived and patient-derived xenografts and reduced tumor burden accompanied with extended survival in a genetically-engineered mouse (GEM) model of SCLC without causing significant thrombocytopenia or other normal tissue injury. Thus, these preclinical findings lay a strong foundation for future clinical studies to test DT2216+mTOR inhibitor combination in a subset of SCLC patients whose tumors are co-driven by BCL-XL and MCL-1.

cancer biology↗

bioRxiv2022

Targeted degradation of microtubule-associated protein tau using an engineered nanobody-E3 ubiquitin ligase adapter fusion

Reducing the level of microtubule-associated protein tau has recently emerged as a promising therapeutic approach for a range of neurodegenerative diseases. Among the various approaches, targeted protein degradation provides a reversible means to rapidly reduce and specifically target disease-relevant forms of tau. However, in aging cells, the protein turnover activity is generally weakened, reducing the efficacy of protein degradation. A potential solution to this is to harness the nuclear proteasomal activity. The nucleus has a high proteasomal content and the degradation activity remains relatively unaffected even in aged cells. Here we show that an E3 ligase F-box domain from the nuclear protein human speckle type BTB/POZ protein (SPOP) is effective in degrading the microtubule-associated protein tau in primary mouse hippocampal neurons. Using EGFP-tagged tau and a GFP-binding nanobody fused to SPOP, we found that the native nuclear localization signal in SPOP causes nuclear sequestration of the target protein. However, degradation of the sequestered target proteins is incomplete, resulting in nuclear accumulation. Replacing the native SPOP nuclear localization signal (NLS) with variants having altered nuclear localization efficiency dramatically affects in the degree of nuclear accumulation of the target protein. Interestingly, nanobody-SPOP with no NLS was more efficient than that with a NLS in reducing overall tau level, causing an approximately 50% reduction in ectopically expressed human tau in mouse neurons. These results show the potential for harnessing the nuclear proteasomal activity for targeted tau degradation in cells and demonstrate a new modality of regulating intracellular protein degradation.

bioengineering↗

bioRxiv2022

Computational Prediction of Synthetic Circuit Function Across Growth Conditions

A challenge in the design and construction of synthetic genetic circuits is that they will operate within biological systems that have noisy and changing parameter regimes that are largely unmeasurable. The outcome is that these circuits do not operate within design specifications or have a narrow operational envelope in which they can function. This behavior is often observed as a lack of reproducibility in function from day to day or lab to lab. Moreover, this narrow range of operating conditions does not promote reproducible circuit function in deployments where environmental conditions for the chassis are changing, as environmental changes can affect the parameter space in which the circuit is operating. Here we describe a computational method for assessing the robustness of circuit function across broad parameter regions. Previously designed circuits are assessed by this computational method and then circuit performance is measured across multiple growth conditions in budding yeast. The computational predictions are correlated with experimental findings, suggesting that the approach has predictive value for assessing the robustness of a circuit design.

synthetic biology↗

bioRxiv2022

Robustness and reproducibility of simple and complex synthetic logic circuit designs using a DBTL loop

Computational tools addressing various components of design-build-test-learn loops (DBTL) for the construction of synthetic genetic networks exist, but do not generally cover the entire DBTL loop. This manuscript introduces an end-to-end sequence of tools that together form a DBTL loop called DART (Design Assemble Round Trip). DART provides rational selection and refinement of genetic parts to construct and test a circuit. Computational support for experimental process, metadata management, standardized data collection, and reproducible data analysis is provided via the previously published Round Trip (RT) test-learn loop. The primary focus of this work is on the Design Assemble (DA) part of the tool chain, which improves on previous techniques by screening up to thousands of network topologies for robust performance using a novel robustness score derived from dynamical behavior based on circuit topology only. In addition, novel experimental support software is introduced for the assembly of genetic circuits. A complete design-through-analysis sequence is presented using several OR and NOR circuit designs, with and without structural redundancy, that are implemented in budding yeast. The execution of DART tested the predictions of the design tools, specifically with regard to robust and reproducible performance under different experimental conditions. The data analysis depended on a novel application of machine learning techniques to segment bimodal flow cytometry distributions. Evidence is presented that, in some cases, a more complex build may impart more robustness and reproducibility across experimental conditions.

synthetic biology↗

bioRxiv2022

A cell-permeant nanobody-based degrader that induces fetal hemoglobin

Proximity-based strategies to degrade proteins have enormous therapeutic potential in medicine, but the technologies are limited to proteins for which small molecule ligands exist. The identification of such ligands for therapeutically relevant but "undruggable" proteins remains challenging. Herein, we employed yeast surface display of synthetic nanobodies to identify a protein ligand selective for BCL11A, a critical repressor of fetal globin gene transcription. Fusion of the nanobody to a cell-permeant miniature protein and an E3 adaptor creates a degrader that depletes cellular BCL11A in erythroid precursor cells, thereby inducing the expression of fetal hemoglobin, a modifier of clinical severity of sickle cell disease and {beta}-thalassemia. This work establishes a new paradigm for the targeted degradation of previously intractable proteins using cell-permeant nanobody-based degraders. One sentence summaryA cell-permeant, protein-based degrader is used for the induction of fetal hemoglobin.

biochemistry↗

bioRxiv2022

Highly-Automated, High-Throughput Replication of Yeast-based Logic Circuit Design Assessments

We describe an experimental campaign that replicated the performance assessment of logic gates engineered into cells of S. cerevisiae by Gander, et al. Our experimental campaign used a novel high throughput experimentation framework developed under DARPAs Synergistic Discovery and Design (SD2) program: a remote robotic lab at Strateos executed a parameterized experimental protocol. Using this protocol and robotic execution, we generated two orders of magnitude more flow cytometry data than the original experiments. We discuss our results, which largely, but not completely, agree with the original report, and make some remarks about lessons learned.

synthetic biology↗

bioRxiv2022

A Microfluidic Chip for LAMP-based Multiplex Detection of Pathogen

Early diagnosis of bacterial causing the disease is important for treatment of patent and preventing the spread of pathogen. Utilizing of POCT devices to detect the pathogens on-site will accelerate the diagnosis of infectious disease. By using loop-mediated-amplification, we developed a microfluidic chip for multiplex detection of three bacterial, where the samples were driven by negative pressure were loaded quickly. The performance of the device was preliminarily evaluated. The specificities of the detections were demonstrated. And the LOD for Escherichia coli, Staphylococcus aureus and Pseudomonas aeruginosa were measured as 17.15, 5.67 and 16.47 ng/L, respectively. The results demonstrated the feasibility of the method.

microbiology↗

bioRxiv2022

Microfluidic chemostatic bioreactor for high-throughput screening and sustainable co-harvesting of biomass and biodiesel in microalgae

As a renewable and sustainable source for energy, environment, and biomedical applications, microalgae and microalgal biodiesel have attracted great attention. However, their applications are confined due to the cost-efficiency of microalgal mass production. One-step strategy and continuous culturing systems could be solutions. However, current studies for optimization throughout microalgae-based biofuel production pipelines are generally derived from the batch culture process. Better tools are needed to study algal growth kinetics in continuous systems. A microfluidics chemostatic bioreactor array was presented, providing low-adhesion cultivation for algae in the gas, nutrition, and temperature (GNT) well-controlled environment with high throughput. The chip wasused to mimic the continuous culture environment of bioreactors. It allowed simultaneously studying of 8x8 different chemostatic conditions on algal growth and oil production in parallel on a 7x7 cm2 footprint. On-chip experiments of batch and continuous cultures of Chlorella. sp. were performed to study growth and lipid accumulation under different nitrogen concentrations. The results demonstrated that microalgal cultures can be regulated to grow and accumulate lipids concurrently, thus enhancing lipid productivity in one step. The developed on-chip culturing condition screening, which was more suitable for continuous bioreactor, was achieved at a half shorter time, 64-times higher throughput, and less reagent consumption. It could be used to establish chemostat cultures in continuous bioreactors which can dramatically accelerate the development of renewable and sustainable algal for CO2 fixation and biosynthesis and related systems for advanced sustainable energy, food, pharmacy, and agriculture with enormous social and ecological benefits. TEASERSustainable microfluidic bioreactor for 64 times higher-throughput screening CO2 fixation and biomass and biodiesel production in microalgae.

bioengineering↗

bioRxiv2022

A microfluidic droplet array demonstrating high-throughput screening in individual lipid-producing microalgae

Microalgae are a group of photoautotrophic microorganisms which could use carbon dioxide for autosynthesis. They have been envisioned as one of the most prospective feedstock for renewable oil. However, great endeavors will still be needed to increase their economic feasibility; the screening of competitive species and suitable culture conditions are such issues. To greatly accelerate these rather laborious steps and also improve their experimental lump-sum-manner, we developed a microfluidic droplet-based 2x103 resolution "identification card", which allowed high throughput real-time monitoring of individual algae among population. A novel fluid-blocking-based droplet generating and trapping performance were integrated in the platform which made it excellent in operational simplicity, rapidity and stability and full of the potentials in single-cell-isolation/screening. The developed platform was successfully used to screen three unicellular algae, namely, Isochrysis zhanjiangensis, Platymonas subcordiformis and Platymonas helgolandica var. tsingtaoensis. In situ bioassays of the lipid accumulation and cell proliferation at single cell level for interspecies comparison were possible. Nitrogen stress condition can be indentified that induce positive-skewed frequency distribution of lipid content.

bioengineering↗

bioRxiv2022

High-efficiency HDR in immortalized cell lines by crude rAAV mediated donor template delivery

Owing to low efficiency of homology-directed repair (HDR), precise knock-in (KI) of large DNA fragments is a challenge in genome editing. High-efficiency HDR has been reported for primary cells in preclinical gene therapy by combining CRISPR/Cas9 mediated induction of double-strand breaks (DSB) with delivery of a single-stranded DNA HDR-donor-template via highly purified recombinant adeno-associated virus (rAAV). Due in part to the labor and expense of rAAV particle purification, rAAV-mediated HDR-template delivery has been underutilized used to generate large KIs in cultured cell lines. Here, we report application of crude preparations of rAAV to deliver HDR-templates for the KI of large [~]2kb fragments at various genomic loci in several -human as well as mouse cell lines at high efficiency. Our approach should facilitate experiments necessitating KI of large DNA fragments to tag endogenous loci for visualization and/or conditional protein degradation.

molecular biology↗

bioRxiv2022

Identification of Piperlongumine (PL) as an new E3 ligase ligand to induce targeted protein degradation

PROteolysis Targeting Chimeras (PROTACs) are bifunctional molecules that degrade target proteins through recruiting E3 ligases. However, their application is limited in part because few E3 ligases can be recruited by known E3 ligase ligands. In this study, we identified piperlongumine (PL), a natural product, as a covalent E3 ligase recruiter, which induces CDK9 degradation when it is conjugated with SNS-032, a CDK9 inhibitor. The lead conjugate 955 can potently degrade CDK9 in a ubiquitin-proteasome-dependent manner and is much more potent than SNS-032 against various tumor cells in vitro. Mechanistically, we identified KEAP1 as the E3 ligase recruited by 955 to degrade CDK9 through a TurboID-based proteomics study, which was further confirmed by KEAP1 knockout and the nanoBRET ternary complex formation assay. In addition, PL-Ceritinib conjugate can degrade EML4-ALK fusion oncoprotein, suggesting that PL may have a broader application as a covalent E3 ligase ligand in targeted protein degradation.

cancer biology↗

bioRxiv2021

DyScore: A Boosting Scoring Method with Dynamic Properties for Identifying True Binders and Non-binders in Structure-based Drug Discovery

The accurate prediction of protein-ligand binding affinity is critical for the success of computer-aided drug discovery. However, the accuracy of current scoring functions is usually unsatisfactory due to their rough approximation or sometimes even omittance of many factors involved in protein-ligand binding. For instance, the intrinsic dynamic of the protein-ligand binding state is usually disregarded in scoring function because these rapid binding affinity prediction approaches are only based on a representative complex structure of the protein and ligand in the binding state. That is, the dynamic protein-ligand binding complex ensembles are simplified as a static snapshot in calculation. In this study, two novel features were proposed for characterizing the dynamic properties of protein-ligand binding based on the static structure of the complex, which is expected to be a valuable complement to the current scoring functions. The two features demonstrate the geometry-shape matching between a protein and a ligand as well as the dynamic stability of protein-ligand binding. We further combined these two novel features with several classical scoring functions to develop a binary classification model called DyScore that uses the Extreme Gradient Boosting algorithm to classify compound poses as binders or non-binders. We have found that DyScore achieves state-of-the-art performance in distinguishing active and decoy ligands on both enhanced DUD dataset and external test sets with both proposed novel features showing significant contributions to the improved performance. Especially, DyScore exhibits superior performance on early recognition, a crucial requirement for success in virtual screening and de novo drug design. The standalone version of DyScore and Dyscore-MF are freely available to all at: https://github.com/YanjunLi-CS/dyscore Key PointsO_LITwo novel binding features were proposed for characterizing the dynamic properties of protein-ligand binding only based on a static snapshot of complex. C_LIO_LIBased on the XGBoost machine learning method, the DyScore recognition model was proposed to accurately classify compound binding poses as binders or non-binders. DyScore consistently outperforms all the state-of-the-art published models on three different metrics by a large margin. C_LIO_LIDyScore showed superior performance in early recognition with an average of 73.3% success rate for the top three ranked compounds for each protein target. C_LIO_LIThe standalone version of DyScore and DyScore-MF are freely available to all at: https://github.com/YanjunLi-CS/dyscore C_LI TOC O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=88 SRC="FIGDIR/small/465921v2_ufig1.gif" ALT="Figure 1"> View larger version (31K): org.highwire.dtl.DTLVardef@1e96bborg.highwire.dtl.DTLVardef@3a0a8corg.highwire.dtl.DTLVardef@8a6f7eorg.highwire.dtl.DTLVardef@9dfa19_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2021

Embeddings of genomic region sets capture rich biological associations in lower dimensions

MotivationGenomic region sets summarize functional genomics data and define locations of interest in the genome such as regulatory regions or transcription factor binding sites. The number of publicly available region sets has increased dramatically, leading to challenges in data analysis. ResultsWe propose a new method to represent genomic region sets as vectors, or embeddings, using an adapted word2vec approach. We compared our approach to two simpler methods based on interval unions or term frequency-inverse document frequency and evaluated the methods in three ways: First, by classifying the cell line, antibody, or tissue type of the region set; second, by assessing whether similarity among embeddings can reflect simulated random perturbations of genomic regions; and third, by testing robustness of the proposed representations to different signal thresholds for calling peaks. Our word2vec-based region set embeddings reduce dimensionality from more than a hundred thousand to 100 without significant loss in classification performance. The vector representation could identify cell line, antibody, and tissue type with over 90% accuracy. We also found that the vectors could quantitatively summarize simulated random perturbations to region sets and are more robust to subsampling the data derived from different peak calling thresholds. Our evaluations demonstrate that the vectors retain useful biological information in relatively lower-dimensional spaces. We propose that vector representation of region sets is a promising approach for efficient analysis of genomic region data. Availabilityhttps://github.com/databio/regionset-embedding

bioinformatics↗

bioRxiv2021

Prediction of Whole-Cell Transcriptional Response with Machine Learning

Applications in synthetic and systems biology can benefit from measuring whole-cell response to biochemical perturbations. Execution of experiments to cover all possible combinations of perturbations is infeasible. In this paper, we present the host response model (HRM), a machine learning approach that takes the cell response to single perturbations as the input and predicts the whole cell transcriptional response to the combination of inducers. We find that the HRM is able to qualitatively predict the directionality of dysregulation to a combination of inducers with an accuracy of >90% using data from single inducers. We further find that the use of known prior, known cell regulatory networks doubles the predictive performance of the HRM (an R2 from 0.3 to 0.65). This tool will significantly reduce the number of high-throughput sequencing experiments that need to be run to characterize the transcriptional impact of the combination of perturbations on the host.

bioinformatics↗