Biology subjects

Lu, C.

Publications and source records attributed to Lu, C..

At least 109 records · Page 6Linked to original sources

bioRxiv2023

DeepHeme: A generalizable, bone marrow classifierwith hematopathologist-level performance

Morphology-based classification of cells in the bone marrow aspirate (BMA) is a key step in the diagnosis and management of hematologic malignancies. However, it is time-intensive and must be performed by expert hematopathologists and laboratory professionals. We curated a large, high-quality dataset of 41,595 hematopathologist consensus-annotated single-cell images extracted from BMA whole slide images (WSIs) containing 23 morphologic classes from the clinical archives of the University of California, San Francisco. We trained a convolutional neural network, DeepHeme, to classify images in this dataset, achieving a mean area under the curve (AUC) of 0.99. DeepHeme was then externally validated on WSIs from Memorial Sloan Kettering Cancer Center, with a similar AUC of 0.98, demonstrating robust generalization. When compared to individual hematopathologists from three different top academic medical centers, the algorithm outperformed all three. Finally, DeepHeme reliably identified cell states such as mitosis, paving the way for image-based quantification of mitotic index in a cell-specific manner, which may have important clinical applications.

bioinformatics↗

bioRxiv2023

Prediction and Design of Protease Enzyme Specificity Using a Structure-Aware Graph Convolutional Network

Site-specific proteolysis by the enzymatic cleavage of small linear sequence motifs is a key post-translational modification involved in physiology and disease. The ability to robustly and rapidly predict protease substrate specificity would also enable targeted proteolytic cleavage - editing - of a target protein by designed proteases. Current methods for predicting protease specificity are limited to sequence pattern recognition in experimentally-derived cleavage data obtained for libraries of potential substrates and generated separately for each protease variant. We reasoned that a more semantically rich and robust model of protease specificity could be developed by incorporating the three-dimensional structure and energetics of molecular interactions between protease and substrates into machine learning workflows. We present Protein Graph Convolutional Network (PGCN), which develops a physically-grounded, structure-based molecular interaction graph representation that describes molecular topology and interaction energetics to predict enzyme specificity. We show that PGCN accurately predicts the specificity landscapes of several variants of two model proteases: the NS3/4 protease from the Hepatitis C virus (HCV) and the Tobacco Etch Virus (TEV) proteases. Node and edge ablation tests identified key graph elements for specificity prediction, some of which are consistent with known biochemical constraints for protease:substrate recognition. We used a pre-trained PGCN model to guide the design of TEV protease libraries for cleaving two non-canonical substrates, and found good agreement with experimental cleavage results. Importantly, the model can accurately assess designs featuring diversity at positions not present in the training data. The described methodology should enable the structure-based prediction of specificity landscapes of a wide variety of proteases and the construction of tailor-made protease editors for site-selectively and irreversibly modifying chosen target proteins.

bioinformatics↗

bioRxiv2023

Changes in physicochemical properties and microbial community succession during leaf stacking fermentation

Leaf stacking fermentation involves enzymatic actions of many microorganisms and is an efficient and environmentally benign process for degrading macromolecular organic compounds. We investigated the dynamics of metabolite profiles, bacterial and fungal communities and their interactions during fermentation using cigar leaves from three geographic regions. The results showed that the contents of total sugar, reducing sugar, starch, cellulose, lignin, pectin, polyphenol and protein in cigar tobacco leaves was significantly decreased during fermentation. Notably, the furfural, neophytadiene, pyridine, benzyl alcohol, geranyl acetone, 2-butanone, 3-hydroxy-, N-hexanal, isoamylalcohol and 2,3-pentanedione were important features volatile aroma compounds during fermentation. The -diversity of fungi and bacteria initially increased and then decreased during fermentation. The microbial diversity was influenced by fermentation stages and growing locations, in which the fermentation stages had greater impacts on the microbial diversity than locations. Microbiome profiling had identified several core bacteria including Sphingomonas, Bacillus, Staphylococcus, Pseudomonas, Ralstonia, Massilia and Fibrobacter. Fungal biomarkers included Aspergillus, Penicillium, Fusarium, Cladosporium and Trichomonascus. Interestingly, the molecular ecological networks showed that the core taxa had significant correlations with metabolic enzymes and physicochemical properties; bacteria and fungi jointly participated in the carbohydrate and nitrogen compound degrading and volatile aroma compound chemosynthesis processes during fermentation. These studies provide insights into the coupling of material conversion and microbial community succession during leaf fermentation. IMPORTANCECigar tobacco leaves are still crude after air curing. Fermentation can eliminate the pungent and bitter taste of the leaves and promote the accumulations of aromatic compounds. Hitherto, there has been no systematic study on the material conversion, metabolic enzymes and microbial community changes and their interactions in fermented leaves. The significance of our research is in identifying the potential metabolic cooperation related to microbial community, which will greatly enhance our understanding of microorganism to convert macromolecular organic to small molecules during the leaf stacking fermentation process.

microbiology↗

bioRxiv2023

Hypothesis-free phenotype prediction within a genetics-first framework

Cohort-wide sequencing studies have revealed that the largest category of variants is those deemed rare, even for the subset located in coding regions (99% of known coding variants are seen in less than 1% of the population1-3). Our understanding of how rare genetic variants influence disease and organism-level phenotypes has achieved limited progress, partly explained by the intrinsic difficulty in statistically evaluating the biological significance of rare events. Here we show that discoveries can instead be made through a knowledge-based approach using protein domains and ontologies (function and phenotype) that considers all coding variants regardless of allele frequency. We describe an ab initio, genetics-first method making molecular knowledge-based interpretations for exome-wide non-synonymous variants for phenotypes at the organism and cellular level. By using this reverse approach, we identify plausible novel genetic causes for developmental disorders that have eluded other established methods and present novel molecular hypotheses for the causal genetics of 40 phenotypes generated from a direct-to-consumer genotype cohort. This system offers a chance to extract further discovery from genetic data after standard tools have been applied.

genomics↗

bioRxiv2023

DPHL v2: An updated and comprehensive DIA pan-human assay library for quantifying more than 14,000 proteins.

A comprehensive pan-human spectral library is critical for biomarker discovery using mass spectrometry (MS)-based proteomics. DPHL v1, a previous pan-human library built from 1096 data-dependent acquisition (DDA) MS data of 16 human tissue types, allows quantifying 10,943 proteins. However, a major limitation of DPHL v1 is the lack of semi-tryptic peptides and protein isoforms, which are abundant in clinical specimens. Here, we generated DPHL v2 from 1608 DDA-MS data acquired using Orbitrap mass spectrometers. The data included 586 DDA-MS newly acquired from 17 tissue types, while 1022 files were derived from DPHL v1. DPHL v2 thus comprises data from 24 sample types, including several cancer types (lung, breast, kidney, and prostate cancer, among others). We generated four variants of DPHL v2 to include semi-tryptic peptides and protein isoforms. DPHL v2 was then applied to a publicly available colorectal cancer dataset with 286 DIA-MS files. The numbers of identified and significantly dysregulated proteins increased by at least 21.7% and 14.2%, respectively, compared with DPHL v1. Our findings show that the increased human proteome coverage of DPHL v2 provides larger pools of potential protein biomarkers.

bioinformatics↗

bioRxiv2022

Heterotrophic bacteria drive sulfide oxidation in coastal sediments

Sulfate reduction and sulfur oxidation are very active in coastal sediments. They shape the biogeochemistry and microbial ecology at hot places of organic matter metabolism. Different from the well-studied sulfate reduction, sulfur oxidation in coastal sediments is still full of questions. Herein, we investigated the distribution of reduced sulfur compounds in differently layers of coastal sediments at the Yellow sea and found that sulfide (H2S), sulfane sulfur (S0), and thiosulfate mainly accumulated in anaerobic sediments and were mostly oxidized in anoxic and oxic interface in the sediments and the sea water. Bacterial community analysis indicated that heterotrophic bacteria are dominating species in surface sediments and sea water. Metagenome analysis showed that two sulfur-oxidizing genes encoding sulfide:quinone oxidoreductases (SQR) and persufide dioxygenases (PDO), were sharply more abundant than other sulfur-oxidizing genes in the coastal sediments. Since members of the marine Roseobacter clade were dominant in coastal waters and sediments, we studied the sulfur oxidation pathway in the Roseobacter Ruegeria pomeroyi DSS-3 and found that sulfide:quinone oxidoreductase, persulfide dioxygenase, and sulfite-oxidizing enzyme were the main enzymes for the oxidation of H2S, zerovalent sulfur, and sulfite/thiosulfate. This study, for the first time, clarified the dominating function of heterotrophic bacteria in sulfur oxidation in the coastal sediments and sea water. IMPORTANCECoastal sediments are the most productive ecosystems. We performed the microbial community diversity and metagenomic analysis of seawater and coastal sediments of the Yellow Sea and explored the sulfur oxidation process in them. We found that heterotrophic bacteria are dominating species in surface sediments and sea water, sulfide and sulfane sulfur were mostly oxidized in surface sediments, and the genes encoding SQR, PDO, and SOE are abundant. Using Ruegeria pomeroyi DSS-3 as the model strain, we studied how these enzymes cooperate to oxidize H2S to sulfate. Thus, this research revealed the critical role of heterotrophic bacteria in sulfur oxidation in coastal sediments and sea water.

microbiology↗

bioRxiv2022

Epigenomic tomography for probing spatially-defined molecular state in the brain

Spatially-resolved genomic profiling is critical for understanding biology in a spatially ordered organ such as mammalian brain. Single-cell spatial genomic assays were developed recently for this purpose but they remain costly and labor-intensive for examining brain tissues across substantial dimensions and surveying a collection of brain samples. Here we demonstrate a new approach, brain epigenomic tomography, that maps spatial epigenomic variations at the scale of centimeters. We profiled neuronal and glial fractions of mouse neocortex slices with 0.5 mm thickness using a low-input technology. H3K27me3 or H3K27ac features across these slices were grouped into clusters based on their spatial variation patterns. Our approach reveals striking dynamics in frontal and caudal cortex due to kainic acid-induced seizure, linked with transmembrane ion transporter, exocytosis of synaptic vesicles, and secretion of neurotransmitter. Epigenomic tomography provides a powerful and cost-effective tool for profiling and discerning brain samples based on their spatial epigenomes.

genomics↗

bioRxiv2022

Smart squirrels use a mortise-tenon structure to fix nuts on understory twigs

Squirrels of temperate zones commonly store nuts or seeds under leaf litter, in hollow logs or even in holes in the ground; however, in humid rainforests, it is an evolutionary challenge for squirrels to hang elliptical or oblate nuts securely on trees to minimize germination or fungal infection. Here, we report a unique behaviour used by two species of small flying squirrels (Hylopetes phayrei electilis (G. M. Allen, 1925) and Hylopetes alboniger (Hodgson, 1870)) to cache nuts on Hainan Island. Squirrels intentionally carved grooves encircling ellipsoid nuts or distributed on the bottom of oblate nuts and used these grooves to fix nuts tightly between small twigs 0.1-0.6 cm in diameter and connected at angles of 25-40{degrees}. The resulting structures were similar to the mortise-tenon joint of ancient Chinese architecture. Cache sites were on small plants located 10-25 m away from the closest potentially nut-producing tree, a behaviour that likely deters discovery and consumption of the nuts by other animals. Their ability to shape individual nuts to store them more securely suggests it is a smart behavior, which is likely formed by the adaption to life in humid tropical rainforests to guarantee food supply and impacts the distribution of tree species.

ecology↗

bioRxiv2022

Spatial Localization Ability of Planarians Identified Through the Light Maze Paradigm

Spatial localization ability is crucial for free-living animals to fit the environment. As shown by previous studies, planarians can be conditioned to discriminate directions. However, due to their simplicity and primitiveness, they had never been considered to have true spatial localization ability to retrieve locations of objects and places in the environment. Here, we introduce a light maze training paradigm to demonstrate that a planarian worm can navigate to a former recognized place from the start point, even if the worm is transferred into a newly produced maze. This finding identifies the spatial localization ability of planarians for the first time, which provides clues for the evolution of spatial learning. Since the planarians have a primitive brain with simple structures, this paradigm can also provide a simplified model for a detailed investigation of spatial learning.

animal behavior and cognition↗

bioRxiv2022

Exploring the genome and protein space of viruses

Recent metagenomic studies have identified a vast number of viruses. However, the systematic assessment of the true genetic diversity of the whole virus community on our planet remains to be investigated. Here, we explored the genome and protein space of viruses by simulating the process of virus discovery in viral metagenomic studies. Among multiple functions, the power function was found to best fit the increasing trends of virus diversity and was therefore used to predict the genetic space of viruses. The estimate suggests that there are at least 8.23e+08 viral Operational Taxonomic Units (vOTUs) and 1.62e+09 viral protein clusters on Earth when assuming the saturation of the virus genetic space, taking into account the balance of costs and the identification of novel viruses. Its noteworthy that less than 3% of the viral genetic diversity has been uncovered thus far, emphasizing the vastness of the unexplored viral landscape. To saturate the genetic space, a total of 3.08e+08 samples would be required. Analysis of viral genetic diversity by ecosystem yielded estimates consistent with those mentioned above. Furthermore, the estimate of the virus genetic space remained robust when accounting for the redundancy of sampling, sampling time, sequencing platform, and parameters used for protein clustering. This study provides a guide for future sequencing efforts in virus discovery and contributes to a better understanding of viral diversity in nature.

microbiology↗

bioRxiv2022

Optimization of genomewide CRISPR screens using AsCas12a and multi-guide arrays

Genomewide loss-of-function (LOF) screening using CRISPR (clustered regularly interspaced short palindromic repeats) has facilitated the discovery of novel gene functions across diverse physiological and pathophysiological systems. A challenge with conventional genomewide CRISPR/Cas9 libraries is the unwieldy size (60,000 to 120,000 constructs), which is resource intensive and prohibitive in some experimental contexts. One solution to streamlining CRISPR screening is by multiplexing two or more guides per gene on a single construct, which enables functional redundancy to compensate for suboptimal gene knockout by individual guides. In this regard, AsCas12a (Cpf1) and its derivatives, e.g., enhanced AsCas12a (enAsCas12a), have enabled multiplexed guide arrays to be specifically and efficiently processed for genome editing. Prior studies have established that multiplexed CRISPR/Cas12a libraries perform comparably to the larger equivalent CRISPR/Cas9 libraries, yet the most efficient CRISPR/Cas12a library design remains unresolved. Here, we demonstrate that CRISPR/Cas12a genomewide LOF screening performed optimally with three guides arrayed per gene construct and could be adapted to robotic cell culture without noticeable differences in screen performance. Thus, the conclusions from this study provide novel insight to streamlining genomewide LOF screening using CRISPR/Cas12a and robotic cell culture.

genetics↗

bioRxiv2022

Effect of F. mosseae on the expression of genes and proteins about sulfur nutrition in the continuous crop soybean roots

Soybean is a sulfur-loving oilseed crop, and continuous cropping can lead to soil sulfur deficiency, which can inhibit the growth and quality of soybean. This experiment used transcriptomic and proteomic sequencing techniques to analyse the changes in the expression of functional genes and related proteins in the root system of continuously cropped soybean and to reveal the molecular mechanism of F. mosseae inoculation on the soybean root system in response to sulfur nutrient supply at the molecular level. It was thus demonstrated that F. mosseae could enhance the uptake and transport of soil sulfur in continuously cropped soybean. This study, therefore, provides a theoretical basis for the application of F. mosseae as a biofertilizer in soybean production on sulfur-deficient soils. One-sentence summaryF. mosseae affects soybean genes and proteins at the transcriptome and proteome levels.

genomics↗

bioRxiv2022

Diving Behavior Reveals Humidity Sensing Ability of Water Deprived Planarians

Humidity sensing ability is crucial to terrestrial animals for fitting the environment. Researchers made great progress in recent study about humidity sensing mechanisms of terrestrial animals. However, it is poorly understood whether humidity sensing exists in aquatic animals. Here, we demonstrate that the aquatic planarians, one of the primitive forerunners of later animals, has the ability of humidity sensing and is capable of using the ability to perceive the water beneath itself from a drought place to seek survival. The behavior we discovered is described as diving because the worms twist its body to break away from the mucus that make them adhere to the drought place and drop into the water. The behavior is triggered by rapidly increasing humidity. This finding suggests that humidity sensing ability exists in the lower aquatic animals, and the ability might be used to seek for water when aquatic animals are facing desiccation. The finding also suggests that survival-seeking and decision-making behavior have appeared in the primitive planarian worms.

animal behavior and cognition↗

bioRxiv2022

Heterologous boost with mRNA vaccines against SARS-CoV-2 Delta/Omicron variants following an inactivated whole-virus vaccine

The coronavirus SARS-CoV-2 has mutated quickly and caused significant global damage. This study characterizes two mRNA vaccines ZSVG-02 (Delta) and ZSVG-02-O (Omicron BA.1), and associating heterologous prime-boost strategy following the prime of a most widely administrated inactivated whole-virus vaccine (BBIBP-CorV). The ZSVG-02-O induces neutralizing antibodies that effectively cross-react with Omicron subvariants following an order of BA.1>BA.2>BA.4/5. In naive animals, ZSVG-02 or ZSVG-02-O induce humoral responses skewed to the vaccines targeting strains, but cellular immune responses cross-react to all variants of concern (VOCs) tested. Following heterologous prime-boost regimes, animals present comparable neutralizing antibody levels and superior protection across all VOCs. Single-boost only generated ancestral and omicron dual-responsive antibodies, probably by "recall" and "reshape" the prime immunity. New Omicron-specific antibody populations, however, appeared only following the second boost with ZSVG-02-O. Overall, our results support a heterologous boost with ZSVG-02-O, providing the best protection against current VOCs in inactivated virus vaccine- primed populations.

molecular biology↗

bioRxiv2022

Conformational change of the Plasmodium TRAP I domain is essential for sporozoite migration and transmission of malaria

Eukaryotic cell adhesion and migration relies on surface adhesins connecting extracellular ligands to the intracellular actin cytoskeleton. Plasmodium sporozoites are transmitted by mosquitoes and rely on adhesion and gliding motility to colonize the salivary glands and to reach the liver after transmission. During gliding the essential sporozoite adhesin TRAP engages actin filaments in the cytoplasm of the parasite., while binding ligands on the substrate through its inserted (I)-domain. Crystal structures of TRAP from different Plasmodium species revealed the I-domain in closed and open conformations. Here, we probe the importance of these two conformational states by generating parasites expressing versions of TRAP with the I-domain stabilized in either the open or closed state with disulfide bonds. Strikingly, both mutations impact sporozoite gliding, mosquito salivary gland entry and transmission. Absence of gliding in sporozoites expressing the open TRAP I-domain could be partly rescued by adding a reducing agent. This suggests that dynamic conformational change is required for ligand binding, gliding motility and organ invasion and hence sporozoite transmission from mosquito to mammal.

cell biology↗

bioRxiv2022

Circulating tumor DNA profiling approach based on in silico background elimination guides patient classification of multiple cancers

Circulating tumor DNA (ctDNA) analysis is increasingly providing a promising minimally invasive alternative to tissue biopsies in precision oncology. However, current methods of ctDNA mutation profiling have limited resolution because of the high background noise and false-positive rate caused by benign variants in plasma cell-free DNA (cfDNA), majorly generated during clonal hematopoiesis. Although personalized parallel white blood cell (WBC) genome sequencing suppresses the noise of clonal hematopoiesis variances observed in ctDNA based liquid biopsy, the system cost and complexity restrict its extensive application in clinical settings. To address this challenge, we describe a Matched WBC Genome sequencing Independent CtDNA profiling (MaGIC) approach, which enables the sensitive detection of recurrent tumor mutant information harbored by ctDNA from a bulk cfDNA background based on hybrid capture cfDNA deep sequencing, in silico background elimination, and a reliable readout measurement by the calculation the number of key mutated exons. Leveraging somatic mutation data from 10163 patients across 24 cancer types obtained from The Cancer Genome Atlas, we confirm that the MaGIC approaches are of ideal performance in prediction of prognosis by tissue biopsy samples of patients across multiple cancers. Meanwhile, MaGIC approaches enable the classification of prostate cancer patients from heathy cohorts by ctDNA sequencing data. We further profiled the ctDNAs of 80 plasma samples from 40 patients with nasopharyngeal carcinoma before and during chemotherapy by MaGIC approaches. The MaGICv2 can predict the chemosensitivity with high accuracy by simply using one liquid biopsy sample of each patient before a stereotypical treatment course. We anticipate that this new approach has the potential utility of ctDNA detection in multiple clinical cancer contexts, thus facilitating precise cancer therapy. TeaserA liquid biopsy analysis method for multifunctional patient classification such as diagnosis and chemosensitivity prediction.

cancer biology↗

bioRxiv2022

Identifying disease-associated circRNAs based on edge-weighted graph attention and heterogeneous graph neural network

MotivationCircular RNAs (circRNAs) with varied biological activities are implicated in pathogenic processes, according to new findings. They are regarded as promising biomarkers for the diagnosis and prognosis due to their structural features. Computational approaches, as opposed to traditional experiments, can identify the circRNA-disease connections at a lower cost. Multi-source pathogenesis data can help to reduce data sparsity and infer probable connections at the system level. The majority of available approaches create a homologous network using multi-source data, but they lose the datas heterogeneity. Effective solutions that make use of the peculiarities of multi-source data are urgently needed. ResultsIn this paper, we propose a model (CDHGNN) based on edge-weighted graph attention and heterogeneous graph neural networks for discovering probable circRNA-disease correlations prediction. The circRNA network, miRNA network, disease network and heterogeneous network are constructed based on the introduced multi-source data on circRNAs, miRNAs, and diseases. The features for each type of node in the network are then extracted using a designed edge-weighted graph attention network model. Using the revised node features, we learn meta-path contextual information and use heterogeneous neural networks to assign attention weights to different types of edges. CDHGNN outperforms state-of-the-art algorithms with comparable accuracy, according to the findings of the trial. Edge-weighted graph attention networks and heterogeneous graph networks have both improved performance significantly. Furthermore, case studies suggest that CDHGNN is capable of identifying particular molecular connections and can be used to investigate pathogenic pathways. Contactjxwang@mail.csu.edu.cn

bioinformatics↗

bioRxiv2022

CRISPR activation rescues abnormalities in SCN2A haploinsufficiency-associated autism spectrum disorder

The majority of autism spectrum disorder (ASD) risk genes are associated with ASD due to haploinsufficiency, where only one gene copy is functional. Here, using SCN2A haploinsufficiency, a major risk factor for ASD, we show that increasing the expression of the existing functional SCN2A allele with CRISPR activation (CRISPRa) can provide a viable therapeutic approach. We first demonstrate therapeutic potential by showing that restoring Scn2a expression in adolescent heterozygous Scn2a conditional knock-in mice rescues electrophysiological deficits associated with Scn2a haploinsufficiency. Next, using an rAAV-CRISPRa based treatment, we restore electrophysiological deficits in both Scn2a heterozygous mice and human stem-cell-derived neurons. Our results provide a novel therapeutic approach for numerous ASD-associated genes and demonstrate that rescue of Scn2a haploinsufficiency, even at adolescent stages, can ameliorate neurodevelopmental phenotypes.

neuroscience↗