Biology subjects

Wu, I.

Publications and source records attributed to Wu, I..

11 recordsLinked to original sources

bioRxiv2026

An interoperable research agent network for scientific discovery

Recent advances in agentic systems have enabled the autonomous execution of research tasks across scientific domains. However, the rapid emergence of specialized scientific agents for areas such as computational pathology, microbiome research, gene editing, materials science, organic chemistry, and drug discovery has created a fragmented ecosystem of scientific capabilities. While these agents often demonstrate strong performance within their respective domains, limited interoperability makes it difficult to combine expertise across platforms and coordinate complex interdisciplinary workflows. Here we introduce GUIA (Guided-research Utilizing Intelligent Agents), an interoperable research-agent network built upon a flexible Agent-to-Agent (A2A) communication architecture. GUIA enables both in-house and third-party agents to collaborate within shared workflows, allowing scientific capabilities to accumulate through the integration of complementary expertise. We evaluated GUIA through four assessments spanning baseline benchmarking, third-party single-agent integration, third-party multi-agent integration, and cross-server agent collaboration. Furthermore, we demonstrate its practical utility through real-world applications involving therapeutic target discovery, drug discovery, and spatial proteomics analysis. Together, our results show that interoperable research-agent networks can coordinate specialized expertise across independently developed systems, providing a scalable framework for expanding scientific capabilities through collaboration.

bioinformatics↗

bioRxiv2026

Organic Components Modulate the Morphology of Respirable Aerovirology Relevant Aerosols

Airborne transmission of pathogens occurs via aerosol particles, whose morphology provides insights into the microenvironments that pathogens experience. Aerosol morphology includes particle size, shape, phase state, and chemical homogeneity, yet systematic studies remain limited. Here, we characterized model bioaerosol morphologies generated from (1) NaCl-organic two-component mixtures, (2) common cell culture media, and (3) artificial respiratory fluids. Particles were collected using a virtual impactor and Andersen cascade impactor and analyzed by scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDX). Results show that organic components modulate the morphology: dipalmitoylphosphatidylcholine (DPPC) promotes organic-inorganic phase separation while proteins prohibit formation of large crystals and leads to better mixing among components. At 30% RH with a drying period of 10 seconds, most aerosols appeared desiccated, though NaCl-glucose, DMEM-complete-media and artificial saliva with mucin remained semi-solid or gel-like. Among all formulations examined EMEM-complete-media and artificial saliva (non mucin) show a size-dependent morphology. Our study demonstrates how chemical composition and size alters surrogate bioaerosol phase (semi-solid or solid) and morphology and provides new insights into the microenvironment of aerosol particles for aerovirology investigations.

biochemistry↗

bioRxiv2025

A single dorsal vagal complex circuit mediates the aversive and anorectic responses to GLP1R agonists

GLP-1 receptor agonists (GLP1RAs) effectively reduce feeding to treat obesity, although nausea and other aversive side effects of these drugs can limit their use. Brainstem circuits that promote satiation and that mediate the physiologic control of body weight can be distinguished from those that cause aversion. It remains unclear whether brainstem Glp1r neurons contribute to the normal regulation of energy balance and whether GLP1RAs control appetite via circuits distinct from those that mediate aversive responses, however. Hence, we defined roles for AP and NTS Glp1r-expressing neurons (APGlp1r and NTSGlp1r neurons, respectively) in the physiologic control of body weight, the GLP1RA-dependent suppression of food intake, and the GLP1RA-mediated stimulation of aversive responses. While silencing non-aversive NTSGlp1r neurons interfered with the physiologic restraint of feeding and body weight, restoring NTSGlp1r neuron Glp1r expression on an otherwise Glp1r-null background failed to enable long-term body weight suppression by GLP1RAs. In contrast, selective Glp1r expression in APGlp1r neurons restored both aversive responses and long-term body weight suppression by GLP1RAs. Thus, while non-aversive NTSGlp1r neurons control physiologic feeding, aversive APGlp1r neurons mediate both the anorectic and weight loss effects of GLP1RAs, dictating the functional inseparability of these pharmacologic GLP1RA responses at a circuit level.

neuroscience↗

bioRxiv2025

Plakophilin-2 Coordinates Energy Metabolism and Contractility in Cardiomyocytes, Revealing Its Roles beyond Desmosomes

Arrhythmogenic right ventricular cardiomyopathy (ARVC) is a familial cardiac disease associated with ventricular arrhythmias and an increased risk of sudden cardiac death. Mutations in the desmosome gene Plakophilin-2, PKP2, lead to reduction in PKP2 protein and collapse of desmosomes that is known to compromise contractility and electrical stability of cardiomyocytes. Our previous studies demonstrated the efficacy of adeno-associated virus 9 (AAV9)-mediated restoration of PKP2 expression in a cardiac specific knock-out mouse model of Pkp2 and revealed profound changes in mRNA signatures of metabolic enzymes that were reversed by the gene replacement approach. In this study, we used PKP2-deficient mouse hearts and human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) to identify changes in steady-state metabolite levels associated with impaired lipid homeostasis, glycolysis, and glucose oxidation. These metabolic phenotypes align with human ARVC metabolic data and reflect an intrinsic impairment of cellular energy metabolism. Here we showed for the first time that these intracellular metabolic defects were associated specifically with poor contractility of cardiomyocytes. AAV9:PKP2 restored contractility, improved electrophysiological properties and Ca2+ transients. In contrast, we observed that treating PKP2-deficient cardiomyocytes pharmacologically with small molecule metabolic enhancers improved contractility but not electrophysiological properties and Ca2+ transients, suggesting differential sensitivity of structure-mediated functions in response to metabolic perturbance. Our study modeled and revealed a direct intracellular connection between compromised PKP2 function and metabolic impairment. We proposed that an increased risk of decoupling energy-responsive contractility from less energy-responsive electrical activities can be a new arrhythmogenic mechanism, potentially responsible for exercise-triggered cardiac adversity in ARVC disease development and progression.

cell biology↗

bioRxiv2025

A Cross-Species Atlas of the Dorsal Vagal Complex Reveals Neural Mediators of Cagrilintide's Effects on Energy Balance

Amylin analogs, including potential anti-obesity therapies like cagrilintide, act on neurons in the brainstem dorsal vagal complex (DVC) that express calcitonin receptors (CALCR). These receptors, often combined with receptor activity-modifying proteins (RAMPs), mediate the suppression of food intake and body weight. To understand the molecular and neural mechanisms of cagrilintide action, we used single-nucleus RNA sequencing to define 89 cell populations across the rat, mouse, and non-human primate caudal brainstem. We then integrated spatial profiling to reveal neuron distribution in the rat DVC. Furthermore, we compared the acute and long-term transcriptional responses to cagrilintide across DVC neurons of rats, which exhibit strong cagrilintide responsiveness, and mice, which respond poorly to cagrilintide over the long term. We found that cagrilintide promoted long-term transcriptional changes, including increased prolactin releasing hormone (Prlh) expression, in the nucleus of the solitary tract (NTS) Calcr/Prlh cells in rats, but not in mice, suggesting the importance of NTS Calcr/Prlh cells for sustained weight loss. Indeed, activating rat area postrema Calcr cells briefly reduced food intake but failed to decrease food intake or body weight over the long term. Overall, these results not only provide a cross-species and spatial atlas of DVC cell populations but also define the molecular and neural mediators of acute and long-term cagrilintide action.

bioinformatics↗

bioRxiv2023

Molecular mechanism of NAD+ binding to the Nudix homology domains of DBC1

DBC1 (deleted in breast cancer 1) is a human nuclear protein that modulates the activities of various proteins. NAD+ (oxidized form of nicotinamide adenine dinucleotide) is thought to potentially bind to the Nudix homology domains (NHDs) of DBC1, thereby regulating DBC1-PARP1 [poly (adenosine diphosphate-ribose) polymerase] interactions, the modulation of which may restore DNA repair to protect against cancer, radiation, and aging. Therefore, our study comprehensively employed methods including NMR (Nuclear Magnetic Resonance), ITC (isothermal titration calorimetry), genetic mutation, and computer biology to thoroughly investigate the molecular mechanism of the binding interaction between NAD+ and its precursor NMN with the NHD domain of DBC1 (DBC1354-396). The results from NMR and ITC indicate that NAD+ likely interacts with DBC1354-396 through hydrogen bonding, with a binding affinity nearly twice that of NMN. The key binding sites are primarily E363 and D372. Molecular Docking further revealed the importance of conventional hydrogen bonds and carbon-hydrogen bonds in the binding process. These findings may lead to a better understanding of how NAD+ regulates the physiological functions of DBC1, thereby offering guiding principles for the development of targeted therapies and drug research focused on tumor diseases associated with DBC1.

biochemistry↗

bioRxiv2022

A Novel Approach to Topological Network Analysis for the Identification of Metrics and Signatures in Non-Small Cell Lung Cancer

Non-small cell lung cancer (NSCLC), the primary histological form of lung cancer, accounts for about 25% - the highest - of all cancer deaths. As NSCLC is often undetected until symptoms appear in the late stages, it is imperative to discover more effective tumor-associated biomarkers for early diagnosis. Topological data analysis is one of the most powerful methodologies applicable to biological networks. However, current studies fail to consider the biological significance of their quantitative methods and utilize popular scoring metrics without verification, leading to low performance. To extract meaningful insights from genomic data, it is essential to understand the relationship between geometric correlations and biological function mechanisms. Through bioinformatics and network analyses, we propose a novel composite selection index, the C-Index, that best captures significant pathways and interactions in gene networks to identify biomarkers with the highest efficiency and accuracy. Furthermore, we establish a 4-gene biomarker signature that serves as a promising therapeutic target for NSCLC and personalized medicine. We designed a Cascading machine learning model to validate both the C-Index and the biomarkers discovered. The methodology proposed for finding top metrics can be applied to effectively select biomarkers and early diagnose many diseases, revolutionizing the approach to topological network research for all cancers.

bioinformatics↗

bioRxiv2020

Accurate Microbiome Sequencing with Synthetic Long Read Sequencing

The microbiome plays a central role in biochemical cycling and nutrient turnover of most ecosystems. Because it can comprise myriad microbial prokaryotes, eukaryotes and viruses, microbiome characterization requires high-throughput sequencing to attain an accurate identification and quantification of such co-existing microbial populations. Short-read next-generation-sequencing (srNGS) revolutionized the study of microbiomes and remains the most widely used approach, yet read lengths spanning only a few of the nine hypervariable regions of the 16S rRNA gene limit phylogenetic resolution leading to misclassification or failure to classify in a high percentage of cases. Here we evaluate a synthetic long-read (SLR) NGS approach for full-length 16S rRNA gene sequencing that is high-throughput, highly accurate and low-cost. The sequencing approach is amenable to highly multiplexed sequencing and provides microbiome sequence data that surpasses existing short and long-read modalities in terms of accuracy and phylogenetic resolution. We validated this commercially-available technology, termed LoopSeq, by characterizing the microbial composition of well-established mock microbiome communities and diverse real-world samples. SLR sequencing revealed differences in aquatic community complexity associated with environmental gradients, resolved species-level community composition of uterine lavage from subjects with histories of misconception and accurately detected strain differences, multiple copies of the 16S rRNA in a single strains genome, as well as low-level contamination in soil cyanobacterial cultures. This approach has implications for widespread adoption of high-resolution, accurate long-read microbiome sequencing as it is generated on popular short read sequencing platforms without the need for additional infrastructure.

microbiology↗

bioRxiv2020

Targeted Transcriptome Analysis using Synthetic Long Read Sequencing Uncovers Isoform Reprograming in the Progression of Colon Cancer

Diversity in human gene expression stems, to a large extent, from splicing exons into multiple mRNA isoforms. Characterization of isoforms requires accurate long-read sequencing. However, read lengths, high error rates, low throughput and large input requirements are some of the challenges that remain to be addressed in sequencing technologies. In this study, we used a barcoding-based synthetic long read (SLR) isoform sequencing approach, LoopSeq, to generate sequencing reads sufficiently long and accurate to identify isoforms using standard short read Illumina sequencers. The method identifies isoforms from control RNA samples with 99.4% accuracy and a 0.01% per-base error rate, exceeding the accuracy reported for other long-read sequencing technologies. Applied to targeted transcriptome sequencing of over 10,000 genes from colon cancers and their metastatic counterparts, LoopSeq revealed large scale isoform redistributions from benign colon mucosa to primary colon cancer and metastatic cancer and identified several novel gene fusion isoforms in the colon cancer samples. Strikingly, our data showed that most single nucleotide variants (SNVs) occurred dominantly in specific isoforms and that some SNVs underwent isoform switching in cancer progression. The ability to use short read sequencers to generate accurate long-read isoform information as the raw unit of transcriptional information holds promise as a new and widely accessible approach in RNA isoform analyses.

genomics↗

bioRxiv2019

A Single-Molecule Long-Read Survey of Human Transcriptomes using LoopSeq Synthetic Long Read Sequencing

State-of-the-art short-read transcriptome sequencing methods employ unique molecular identifier (UMI) to accurately classify and count mRNA transcripts. A fundamental limitation of UMI-based short-read transcriptome sequencing is that each read typically covers a small fraction of the transcript sequence. Efforts to accurately characterize splicing isoforms, arguably the largest source of variation in Human gene expression, using short read sequencing have therefore largely relied on computational predictions of transcript isoforms based on indirect observations. Here we describe a transcript counting, synthetic long read method for sequencing whole transcriptomes using short read sequencing platforms and no additional hardware. The method enables full-length mRNA sequence reconstruction at single-nucleotide resolutions with high-throughput, low error rates and UMI based transcript counting using any Illumina sequencer. We describe results from whole transcriptome sequencing from total RNA extracted from 3 human tissue samples: brain, liver, and blood. Reconstructed transcript sequences are characterized and annotated using SQANTI, an analysis pipeline for assessing the sequence quality of long-read transcriptomes. Our results demonstrate that LoopSeq synthetic long-read sequencing can reconstruct contigs up to 3,900nt full-length transcripts using tissue extracted RNA, as well as identify novel splice variants of known junction donors and acceptors.

genomics↗

bioRxiv2017

Resolving the Full Spectrum of Human Genome Variation using Linked-Reads

Large-scale population based analyses coupled with advances in technology have demonstrated that the human genome is more diverse than originally thought. To date, this diversity has largely been uncovered using short read whole genome sequencing. However, standard short-read approaches, used primarily due to accuracy, throughput and costs, fail to give a complete picture of a genome. They struggle to identify large, balanced structural events, cannot access repetitive regions of the genome and fail to resolve the human genome into its two haplotypes. Here we describe an approach that retains long range information while harnessing the advantages of short reads. Starting from only [~]1ng of DNA, we produce barcoded short read libraries. The use of novel informatic approaches allows for the barcoded short reads to be associated with the long molecules of origin producing a novel datatype known as Linked-Reads. This approach allows for simultaneous detection of small and large variants from a single Linked-Read library. We have previously demonstrated the utility of whole genome Linked-Reads (lrWGS) for performing diploid, de novo assembly of individual genomes (Weisenfeld et al. 2017). In this manuscript, we show the advantages of Linked-Reads over standard short read approaches for reference based analysis. We demonstrate the ability of Linked-Reads to reconstruct megabase scale haplotypes and to recover parts of the genome that are typically inaccessible to short reads, including phenotypically important genes such as STRC, SMN1 and SMN2. We demonstrate the ability of both lrWGS and Linked-Read Whole Exome Sequencing (lrWES) to identify complex structural variations, including balanced events, single exon deletions, and single exon duplications. The data presented here show that Linked-Reads provide a scalable approach for comprehensive genome analysis that is not possible using short reads alone.

genomics↗