Biology subjects

Xing, E.

Publications and source records attributed to Xing, E..

15 recordsLinked to original sources

bioRxiv2026

Dehydration-Responsive Left-Handed G-Quadruplexes Expandthe Sequence Repertoire

G-quadruplexes (G4s) are widely studied noncanonical nucleic-acid structures that carry important biological functions. Typically, G4s form from G-rich sequences containing at least four tracts of more than two consecutive guanines, yet a plethora of unconventional G4-forming sequences have been identified, including sequences that fold into G4s only under specific conditions. These G4 states remain poorly resolved. Here we identify dehydration-responsive left-handed G4s (drLHG4s), a class of left-handed G4 (LHG4) structures accessible under dehydrating conditions. Using d(GT)n repeats, we show that left-handed folding depends on repeat length, salt condition and solvent conditions. Systematic mutational analysis reveals a broader but register-constrained sequence space, while high-resolution crystal structures of d-GTGGTGTGTGGTG, a sequence that folded robustly under dehydrating conditions, establish a bulged LHG4 architecture and reveal a distinctive hydration sphere. We further show that selected peptides and a cationic zinc phthalocyanine derivative, ZnTEGQPyrPc, containing four triethylene glycol methyl ether tails can induce LHG4 states. Together, our findings reveal a dehydration-responsive G4 landscape and expand the experimentally accessible sequence space of LHG4 folding.

biophysics↗

bioRxiv2026

Design, Amino-Acid-Derived Analogues of FTY-720 and OSU-2S as PP2A Activators: Structure-Activity Relationships and Identification of Lead Compound XML-3D

Protein phosphatase 2A (PP2A) is a critical regulator of oncogenic signaling pathways, and pharmacological reactivation of PP2A through disruption of inhibitory protein interactions has emerged as a promising anticancer strategy. Although the sphingosine analogue FTY-720 and its derivative OSU-2S can restore PP2A activity, further structural diversification is needed to improve potency and expand structure-activity understanding. Here, we report the design, synthesis, and biological evaluation of three series of amino-acid-derived analogues incorporating tyrosine, homotyrosine, and tryptophan scaffolds. Docking models were used qualitatively to guide exploration of head-group functionality, hydrophobic tail variation, and stereochemical configuration; these models were hypothesis-generating and were not experimentally validated. Modular synthetic routes enabled efficient preparation of structurally diverse compounds from naturally occurring starting materials. Biological evaluation in leukemia and lymphoma cell lines revealed that hydrophobic tail length strongly influenced cytotoxic activity, while tryptophan-derived analogues were generally less potent than tyrosine-based scaffolds. Shortening the hydrophobic tail did not improve activity, contrary to the initial docking-based hypothesis. Stereochemical inversion at the amino alcohol head group yielded a lead compound, XML-3D, which demonstrated approximately two-fold enhanced cytotoxicity relative to FTY-720 and OSU-2S across leukemia cell lines (IC50 = 4.1, 1.8, and 1.5 uM in MV4-11, HL-60, and MEC-1 cells, respectively) and showed broader activity in mantle cell lymphoma lines (IC50 = 530 nM, 970 nM, and 110 nM in Jeko-1, Mino-1, and UPN1 cells). XML-3D induced increased PP2A phosphatase activity in cellular assays. Co-immunoprecipitation and pharmacological inhibition experiments were consistent with modulation of the PP2A/SET regulatory axis; however, direct binding measurements were not performed, and the mechanistic interpretation remains provisional. These findings suggest preliminary structure-activity relationships for amino-acid-derived PP2A modulators and identify stereochemical optimization as a productive direction for future development. Ongoing work aims to define molecular target engagement and refine mechanistic understanding to guide next-generation design.

molecular biology↗

bioRxiv2026

Large Language Model Agent Enables Autonomous Machine Learning Model Building for Biomedicine

Machine learning accelerates biomedical discovery, but creating effective predictive models requires specialized human expertise and demanding manual effort. Researchers must iteratively design pipelines, select architectures, and debug code. This challenge is particularly severe in biomedicine because of the heterogeneous datasets, sparse annotations, and complex evaluation protocols that are common in the domain. We present AIDO.Builder, an agentic artificial intelligence system that fully automates the entire life-cycle of biomedical model development. Provided only with a natural language task description and a target metric, AIDO.Builder autonomously constructs executable training and evaluation pipelines. The system selects suitable modeling strategies, executes experiments, and uses automated feedback-loop to iteratively revise its own code, configurations, and training procedures. It flexibly adapts to new tasks by training specialized models de novo or by using pretrained foundation models to build predictive models through task-appropriate adaptation. We show that across diverse biomedical benchmarks, AIDO. Builder produces highly competitive solutions against human alternatives, while eliminating the manual iteration previously required for robust model development. By automating the translation of raw data into reliable AI models, AIDO.Builder demonstrates how AI itself can be used to accelerate AI for biomedical research.

bioinformatics↗

bioRxiv2025

PertAdapt: Unlocking Single-Cell Foundation Models for Genetic Perturbation Prediction via Condition-Sensitive Adaptation

Single-cell foundation models (FMs) pretrained on massive unlabeled scRNA-seq data show strong potential in predicting transcriptional responses to unseen genetic perturbations. However, existing approaches insufficiently transfer pretrained knowledge and overlook the imbalance between perturbation-sensitive and insensitive genes, yielding only marginal improvements over nonpretrained baselines. To address these limitations, we introduce PertAdapt, a framework that unlocks FMs to accurately predict genetic perturbation effects via integrating a plug-in perturbation adapter and an adaptive loss. The adapter employs a gene-similarity-masked attention mechanism to jointly encode perturbation conditions and contextualized representations of unperturbed cells, enabling more effective knowledge transfer. To better capture differential expression patterns, the adaptive loss dynamically reweights perturbation-sensitive genes relative to global transcriptomic signals. Extensive experiments across seven perturbation datasets, including both single- and double-gene settings, demonstrate that PertAdapt consistently outperforms non-pretrained and FM baselines. Moreover, PertAdapt demonstrates strong capacity for modeling multiplexed gene interactions, generalizing in limited-data regimes, and maintaining robustness across backbone sizes. AvailabilityCode and data are available at https://github.com/BaiDing1234/PertAdapt.

bioinformatics↗

bioRxiv2025

MSACLR: Contrastive Learning of Protein Conformations from MSAs

We propose MSACLR (Multiple Sequence Alignment Contrastive Learning Representation), a two-stage contrastive learning framework that maps MSA space to conformational space. In Stage 1, embeddings are trained to discriminate structural folds across diverse proteins using only MSA information. In Stage 2, embeddings are fine-tuned on subMSAs labeled by their associated predicted structural clusters, enabling discrimination of alternative conformations within the same protein. To enrich training data, we introduce BLOSUM62-guided [1] augmentation, which expands the pool of subMSAs associated with each structural cluster label by introducing sequence-level diversity. Our experiments show that MSACLR embeddings achieve clearer fold-level separation than single-sequence baselines, while fine-tuned embeddings capture conformational variation across scales--from local loop motions to domain motions and fold switching. MSACLR provides a foundation for efficient exploration of MSA space and enables sampling of conformational ensembles, bridging the gap between static structure prediction and dynamic protein behavior.

bioinformatics↗

bioRxiv2025

The pH gradient contributes to persistence in Mycobacterium tuberculosis

Tuberculosis (TB) remains difficult to cure due in part to poorly defined drug-tolerant persister cells formed by Mycobacterium tuberculosis (Mtb), which survive antibiotic treatment without evidence of genetic resistance. To better define this phenotype, we screened 2,336 FDA-approved drugs for compounds that target persistence. Unexpectedly, we identified a strong inducer of drug tolerance -- the antiparasitic niclosamide (NCA), which is known to disrupt proton motive force. In contrast to earlier reports that it harbors promising anti-TB activity, we found that NCA protected Mtb from bactericidal doses of isoniazid, rifampicin, and other standard TB drugs. Investigating further, we showed that disruption of the pH gradient and consequent intracellular acidification is needed to induce tolerance, while disruption of membrane potential is not, and also that protection is tunable by external pH. Transcriptomic analysis of these chemically-induced persister (CIP) cells implicated specific genes in this phenotype, and targeted knockdowns confirmed roles for three genes in either promoting or mitigating the tolerance state. These findings highlight that chemical disruption of the pH gradient is a facile and rapid means to induce drug tolerance, offering a potentially useful tool to probe persister biology in TB and other infectious diseases.

microbiology↗

bioRxiv2025

Discovery of Niclosamide Analogs with Potent Mitochondrial Uncoupling Activity with reduced toxicity

Mitochondrial uncouplers have shown clinical potential across various diseases, including cancer. Niclosamide, an FDA-approved anthelmintic drug, acts as a mild mitochondrial uncoupler and has demonstrated anticancer activity in multiple preclinical cancer models. However, its clinical application remains limited, with some attributing this to poor bioavailability, while the underlying mechanisms are still unclear. Here, we demonstrate that niclosamide exhibits a dose-dependent biphasic effect, promoting uncoupling at low concentration while acting as a mitochondrial inhibitor at high concentration, which could restrict its therapeutic window and limit efficacy. To overcome this challenge, we aimed to develop next-generation mitochondrial uncouplers (MUs) by synthesizing and evaluating novel Niclosamide derivatives with enhanced therapeutic potential. Through structural modifications, we optimized uncoupling activity while reducing inhibitory toxicity, thereby expanding the pharmacological window. Our findings suggest that fine-tuning the molecular structure of mitochondrial uncouplers could provide a safer and more effective metabolic reprogramming strategy for cancer treatment.

pharmacology and toxicology↗

bioRxiv2025

AIDO.Tissue: Spatial Cell-Guided Pretraining for Scalable Spatial Transcriptomics Foundation Model

Single-cell spatial transcriptomics enables high-resolution insights into tissue organization and cell-cell interactions, yet poses significant computational and modeling challenges due to its scale and complexity. Here we introduce AIDO.Tissue, a spatially-informed pretraining framework. The design employs multiple cells as input and an asymmetric encoder-decoder architecture, making it effectively encodes cross-cell dependencies while scaling to large data. Systematic evaluation shows that our method scales with neighboring size and achieves state-of-the-art performance across diverse downstream tasks, including spatial cell type classification, cell niche type prediction and cell density estimation. These results highlight the importance of spatial context in building general-purpose foundation models for tissue-level understanding.

bioinformatics↗

bioRxiv2025

Interpretable adenylation domain specificity prediction using protein language models

Natural products have long been a rich source of diverse and clinically effective drug candidates. Non-ribosomal peptides (NRPs), polyketides (PKs), and NRP-PK hybrids are three classes of natural products that display a broad range of bioactivities, including antibiotic, antifungal, anticancer, and immunosuppressant activities. However, discovering these compounds through traditional bioactivity-guided techniques is costly and time-consuming, often resulting in the rediscovery of known molecules. Consequently, genome mining has emerged as a high-throughput strategy to screen hundreds of thousands of microbial genomes to identify their potential to produce novel natural products. Adenylation domains play a key role in the biosynthesis of NRPs and NRP-PKs by recruiting substrates to incrementally build the final structure. We propose MASPR, a machine learning method that leverages protein language models for accurate and interpretable predictions of A-domain substrate specificities. MASPR demonstrates superior accuracy and generalization over existing methods and is capable of predicting substrates not present in its training data, or zero-shot classification. We use MASPR to develop Seq2Hybrid, an efficient algorithm to predict the structure of hybrid NRP-PK natural products from microbial genomes. Using Seq2Hybrid, we propose putative biosynthetic gene clusters for the orphan natural products Octaminomycin A, Dityromycin, SW-163B, and JBIR-39.

bioinformatics↗

bioRxiv2024

scLong: A Billion-Parameter Foundation Model for Capturing Long-Range Gene Context in Single-Cell Transcriptomics

Single-cell RNA sequencing (scRNA-seq) has revolutionized the study of cellular heterogeneity by providing gene expression data at single-cell resolution, uncovering insights into rare cell populations, cell-cell interactions, and gene regulation. Foundation models pretrained on large-scale scRNA-seq datasets have shown great promise in analyzing such data, but existing approaches are often limited to modeling a small subset of highly expressed genes and lack the integration of external genespecific knowledge. To address these limitations, we present sc-Long, a billion-parameter foundation model pretrained on 48 million cells. sc-Long performs self-attention across the entire set of 28,000 genes in the human genome. This enables the model to capture long-range dependencies between all genes, including lowly expressed ones, which often play critical roles in cellular processes but are typically excluded by existing foundation models. Additionally, sc-Long integrates gene knowledge from the Gene Ontology using a graph convolutional network, enriching its contextual understanding of gene functions and relationships. In extensive evaluations, sc-Long surpasses both stateof-the-art scRNA-seq foundation models and task-specific models across diverse tasks, including predicting transcriptional responses to genetic and chemical perturbations, forecasting cancer drug responses, and inferring gene regulatory networks.

bioinformatics↗

bioRxiv2024

Netupitant Exhibits Potent Activity on Mycobacterium tuberculosis Persisters

In Mycobacterium tuberculosis (Mtb), persisters are genotypically drug-sensitive bacteria that nonetheless survive antibiotic treatment. Persisters represent a significant challenge to shortening TB treatment and preventing relapse, underscoring the need for new therapeutic strategies. In this study, we screened 2,336 FDA-approved compounds to identify agents that enhance the sterilizing activity of standard anti-TB drugs and prevent the regrowth of persisters. Netupitant (NTP), an FDA-approved antiemetic, emerged as a promising candidate with bacteriostatic activity on its own. However, in combination with isoniazid (INH) and rifampicin (RIF), NTP eliminated viable Mtb cells within 7 days, achieving a >6-log reduction in colony-forming units (CFUs) compared to the 2.5-log reduction observed with INH-RIF alone. NTP also demonstrated broad-spectrum efficacy, enhancing the activity of multiple TB drugs, including ethambutol, moxifloxacin, amikacin, and bedaquiline. Notably, NTP retained its potency under hypoxic and caseum-mimicking conditions, both of which are known to enrich for non-replicating, drug-tolerant cells. Interestingly, under hypoxic conditions, NTP demonstrated strong tuberculocidal activity, achieving an approximate 4-log CFU reduction, whereas high-dose INH-RIF was ineffective. Transcriptomic analysis revealed that NTP primarily disrupts cellular bioenergetics, with significant downregulation observed in activities associated with the electron transport chain, oxidative phosphorylation, NADH-ubiquinone oxidoreductase, succinate dehydrogenase, and ATP synthesis. While further studies are required to decipher the mechanism of action and resistance profile of NTP, and to assess its in vivo efficacy, these findings underscore its potential as a promising adjunct to existing TB therapies.

microbiology↗

bioRxiv2024

ProteinAligner: A Multi-modal Pretraining Framework for Protein Foundation Models

Protein foundation models, particularly protein language models, have demonstrated strong success in learning meaningful representations of proteins using transformer architectures pretrained on large-scale protein datasets with self-supervised learning. These representations have been highly effective for downstream tasks such as predicting protein functions and properties. However, most current protein foundation models focus on pretraining with amino acid sequences, often neglecting additional modalities like protein structures and related literature, both of which provide valuable insights. To address this gap, we propose a multi-modal pretraining approach that integrates three key modalities - protein sequences, structures, and literature text. In our framework, the protein sequence modality serves as the anchor, with the other two modalities aligned to it, enhancing the models capacity to capture more comprehensive protein information. ProteinAligner out-performed state-of-the-art protein foundation models in predicting protein functions and properties across diverse down-stream tasks.

bioinformatics↗

bioRxiv2024

AttentionPert: Accurately Modeling Multiplexed Genetic Perturbations with Multi-scale Effects

Genetic perturbations (i.e. knockouts, variants) have laid the foundation for our understanding of many diseases, implicating pathogenic mechanisms and indicating therapeutic targets. However, experimental assays are fundamentally limited in the number of perturbation conditions they can measure. Computational methods can fill this gap by predicting perturbation effects under unseen conditions, but accurately predicting the transcriptional responses of cells to unseen perturbations remains a significant challenge. We address this by developing a novel attention-based neural network, AttentionPert, which accurately predicts gene expression under multiplexed perturbations and generalizes to unseen conditions. AttentionPert integrates global and local effects in a multi-scale model, representing both the non-uniform system-wide impact of the genetic perturbation and the localized disturbance in a network of gene-gene similarities, enhancing its ability to predict nuanced transcriptional responses to both single and multi-gene perturbations. In comprehensive experiments, AttentionPert demonstrates superior performance across multiple datasets outperforming the state-of-theart method in predicting differential gene expressions and revealing novel gene regulations. AttentionPert marks a significant improvement over current methods, particularly in handling the diversity of gene perturbations and in predicting out-of-distribution scenarios.

bioinformatics↗

bioRxiv2020

Supervised Adversarial Alignment of scRNA-seq Data

Dimensionality reduction is an important first step in the analysis of single cell RNA-seq (scRNA-seq) data. In addition to enabling the visualization of the profiled cells, such representations are used by many downstream analyses methods ranging from pseudo-time reconstruction to clustering to alignment of scRNA-seq data from different experiments, platforms, and labs. Both supervised and unsupervised methods have been proposed to reduce the dimension of scRNA-seq. However, all methods to date are sensitive to batch effects. When batches correlate with cell types, as is often the case, their impact can lead to representations that are batch rather than cell type specific. To overcome this we developed a domain adversarial neural network model for learning a reduced dimension representation of scRNA-seq data. The adversarial model tries to simultaneously optimize two objectives. The first is the accuracy of cell type assignment and the second is the inability to distinguish the batch (domain). We tested the method by using the resulting representation to align several different datasets. As we show, by overcoming batch effects our method was able to correctly separate cell types, improving on several prior methods suggested for this task. Analysis of the top features used by the network indicates that by taking the batch impact into account, the reduced representation is much better able to focus on key genes for each cell type.

bioinformatics↗

bioRxiv2018

Removing Confounding Factors Associated Weights in Deep Neural Networks Improves the Prediction Accuracy for Healthcare Applications

The proliferation of healthcare data has brought the opportunities of applying data-driven approaches, such as machine learning methods, to assist diagnosis. Recently, many deep learning methods have been shown with impressive successes in predicting disease status with raw input data. However, the \"black-box\" nature of deep learning and the high-reliability requirement of biomedical applications have created new challenges regarding the existence of confounding factors. In this paper, with a brief argument that inappropriate handling of confounding factors will lead to models sub-optimal performance in real-world applications, we present an efficient method that can remove the influences of confounding factors such as age or gender to improve the across-cohort prediction accuracy of neural networks. One distinct advantage of our method is that it only requires minimal changes of the baseline models architecture so that it can be plugged into most of the existing neu-ral networks. We conduct experiments across CT-scan, MRA, and EEG brain wave with convolutional neural networks and LSTM to verify the efficiency of our method.

bioinformatics↗