Biology subjects

Ye, E.

Publications and source records attributed to Ye, E..

9 recordsLinked to original sources

bioRxiv2026

Ultrafast and Accurate Selection of High-Quality Protein Complex Models from Large-Scale Prediction

In the post-AlphaFold era, advances in protein complex structure prediction have enabled the generation of thousands of candidate models per target, making the accurate and efficient selection of high-quality models from large candidate pools a critical challenge. Existing model-selection approaches commonly employ estimated model accuracy (EMA) methods, which include single-model and consensus-based methods. The former lack cross-model comparative information, whereas the latter often incur high computational costs from pairwise structural alignments and are influenced by candidate-pool quality. Here, we present DeepUMQA-Selection, a two-stage framework combining single-model accuracy estimators for preliminary selection with Ultrafast Shape Recognition (USR)-based structural alignment for consensus selection. Preliminary selection enriches the candidate pool with high-quality models and reduces subsequent structural comparisons, while USR-based structural alignment enables efficient consensus selection without explicit structural superposition. On the retrospective CASP16 QMODE3 benchmark dataset, DeepUMQA-Selection outperformed all participating methods and predictor-derived confidence metrics, as measured by mean Top-5 weighted penalty. The advantage was particularly pronounced for heteromeric targets, with a 13.9% improvement in model-selection performance relative to the best-performing CASP16 participating method. Furthermore, a confidence-assisted strategy enabled the framework to accommodate larger protein complexes while maintaining superior model-selection performance over competing methods, supporting its scalability. Runtime benchmarking showed that the proposed USR-based structural alignment achieved a substantial 317-fold speedup over US-align for candidate pools containing 1,000 models, demonstrating its computational efficiency for large-scale model selection. Our results demonstrate that DeepUMQA-Selection enables accurate and computationally efficient selection of high-quality protein complex models from large-scale candidate pools.

bioinformatics↗

bioRxiv2026

Biologically grounded cell profiling across microscopy modalities

Microscopy-based cell profiling has broad applications in biological discovery, disease characterization, and phenotypic drug screening. Modern microscopy continues to push the limits of resolution, speed, depth and throughput, but better imaging does not automatically lead to better biomedical discovery and translation. A key bottleneck is feature representation: existing features are either handcrafted or learned as black-box embeddings and often lack explicit biological meaning. Here we propose biological grounding as a first principle for cell profiling and implement it through MorphAgent, an AI agent framework in which each quantitative feature originates from a biological hypothesis and is anchored to a cellular structure or process. By integrating biological knowledge, multimodal reasoning and automated validation, MorphAgent makes biologically grounded feature design systematic and scalable. We show biological grounding fundamentally changes the properties of cell profiling features. Across three microscopy modalities, MorphAgent produces compact, expressive and transferable features for drug screening, cell state characterization and disease morphology analysis. In wide-field Cell Painting, a common assay for phenotypic drug screening, it improves perturbation-retrieval mean average precision by 48% over CellProfiler and 20% over DeepProfiler while using substantially fewer dimensions. In confocal mitochondrial imaging, biologically grounded features transfer across independently acquired datasets and imaging resolutions, supporting accurate aging-state classification. In structured illumination microscopy (SIM) imaging of Tau-labeled samples, biologically grounded super resolution features reveal nanoscale morphologies inaccessible at conventional resolution, enhancing the discrimination of disease-associated mutations. Moreover, biologically grounded features provide a hierarchical organization of cellular morphology, enabling multilevel alignment with transcriptomics data and facilitating mechanistic understanding. Biologically grounded cell profiling thus brings AI reasoning closer to biological mechanisms, advancing biomedical discovery and translation.

bioinformatics↗

bioRxiv2026

Activity-based profiling of primary brain cells identifies covalent allosteric modulators of HCN channels

Chemical proteomics can provide global portraits of small molecule-protein interactions in native biological systems. Such ligandability maps have, however, been mostly restricted to readily accessible cell lines and primary immune cells. Here, we describe an activity-based protein profiling (ABPP) strategy for mapping the covalent ligandability of primary brain cells isolated from mice. By investigating sets of stereochemically defined electrophilic small molecules (stereoprobes), we identify liganding events for diverse brain cell proteins, including many with nervous system-enriched expression. In this category were multiple hyperpolarization-activated cyclic nucleotide-gated (HCN) ion channels, which we show are covalently liganded by tryptoline acrylamide stereoprobes at a conserved cysteine in their cyclic nucleotide-binding domain. The stereoprobes were found to block cAMP-dependent shifts in voltage dependence while sparing basal activity of HCN channels. We thus describe an advanced ABPP platform for identifying ligands targeting nervous system-enriched proteins, including chemical probes that modulate HCN channel function in cells.

biochemistry↗

bioRxiv2026

Highly Accurate Estimation of the Fold Accuracy of Protein Structural Models

BackgroundThe function of a protein is intrinsically linked to its three-dimensional fold, and deep learning has revolutionized the field by enabling high-accuracy structure prediction at an unprecedented scale. Nevertheless, the growing deployment of these predictive pipelines in drug discovery and structural biology reveals a critical bottleneck that lies in the lack of independent and rigorous estimation of model accuracy (EMA) methodologies. ResultsHere we present DeepUMQA-Global, a single-model deep learning framework for estimating accuracy of protein structure models. Our method employs a structure-sequence cross-consistency mechanism to evaluate the bidirectional compatibility between the predicted structure and the input sequence, enabling comprehensive characterization of fold accuracy. DeepUMQA-Global outperforms the self-assessment confidence scores of AlphaFold3, achieving improvements of 57.8% in Pearson correlation and 49.0% in Spearman correlation. With respect to the CASP16 retrospective benchmark, DeepUMQA-Global outperforms all single-model accuracy estimation methods that participated in CASP16 and achieves performance comparable to that of the top consensusObased methods. A lightweight consensus strategy built upon DeepUMQA-Global ranks first among all CASP16 participants, surpassing all other methods, including consensus approaches, and highlighting the strength of our method. Remarkably, DeepUMQA-Global demonstrates a strong ability to discriminate between alternative conformational states of proteins, as evidenced in the CASP unique alternative conformation protein complex target and the CoDNaS benchmark. ConclusionsOur results indicate that DeepUMQA-Global can be extended to broader protein modeling tasks, moving beyond static evaluation to offer a foundation for dynamic conformation EMA, where it accurately discriminates alternative conformational states and demonstrates reliable predictive fidelity in model accuracy estimation.

bioinformatics↗

bioRxiv2026

Non-invasive Neuromodulation Targeting Approach by Mapping Stimulations and Lesions That Modify Visual Memory

Therapeutic brain stimulation is believed to target specific networks, but targeting approaches for memory remain debated. For other symptoms, neuromodulation targets have been localized by mapping connectivity of lesions and stimulation sites to specific symptoms. This approach has yielded networks for global memory, but it remains unclear whether it applies to specific types of memory. Here, we mapped connectivity of stimulation sites, lesions, and atrophy patterns associated with different memory types. We included 544 individuals across three datasets: transcranial magnetic stimulation (N=262), penetrating head trauma (N=169), and ischemic stroke (N=113). We identified a network preferentially connected to lesions and stimulation sites specifically associated with changes in visual memory. Of note, the direction of this effect was inverted depending on whether lesions or stimulation occurred at younger age or an older age, consistent with prior results. This age effect was replicated in an independent dataset of patients with preclinical Alzheimers disease (N=1240). To examine neuromodulation targets, we computed electrical field models for potential TMS sites that overlap with the networks derived from each stimulation or lesion dataset; the resulting targets intersected with established targets that demonstrated efficacy for treating memory impairment - precuneus, cortical-hippocampal network, and dorsolateral prefrontal cortex - with peak intersection at medial posterior parietal lobe, angular gyrus, and left anterior middle frontal gyrus, respectively. Future head-to-head clinical trials are needed to systematically compare these proposed neuromodulation targets against each other. One Sentence SummaryNeuromodulation targets for visual memory diverge by age at the time of injury or stimulation.

neuroscience↗

bioRxiv2025

MViewEMA: Efficient Global Accuracy Estimation for Protein Complex Structural Models Using Multi-View Representation Learning

Estimation of model accuracy (EMA) is crucial for the reliable application of protein structure models in biological research. With the rapid advancement of protein structure prediction techniques and the explosive growth of predicted structural data, existing EMA methods struggle to balance computational efficiency with estimation performance. Here, we present MViewEMA, a single-model EMA method that leverages a multi-view representation learning framework to integrate residue-residue interaction features from micro-environment, meso-environment, and macro-environment levels for global accuracy assessment of protein complex models. Benchmark results demonstrate that MViewEMA outperforms state-of-the-art EMA methods in global accuracy assessment, achieving more than a 10-fold improvement in computational efficiency compared to our previous method, DeepUMQA3. This method enables efficient selection of high-quality protein complex models from large-scale structural datasets and achieved top performance in model selection tracks during the CASP16 blind test, demonstrating its potential to enhance the accuracy of complex structure prediction when integrated into modern frameworks such as AlphaFold-Multimer, AlphaFold3, and DiffDock-PP.

bioinformatics↗

bioRxiv2024

Genome-wide CRISPR screen identifies AC9 as a key regulator of ER calcium homeostasis involved in neuronal differentiation.

Endoplasmic reticulum (ER) calcium (Ca2+) homeostasis is essential for maintaining normal cellular physiological functions. Its disturbance is strongly linked to the onset and progression of human diseases, including cancer, developmental defects, and neurodegenerative disorders. The lack of sensitive ratiometric ER Ca2+ indicators, nevertheless, hinders systematic investigation of ER Ca2+ modulators and the underlying mechanisms. Capitalizing on two ultra-sensitive ER Ca2+ indicators and CRISPR-based genome-wide screening, we identified a set of proteins capable of reducing the ER Ca2+ content. Further comparative analysis and qPCR validation pinpointed adenylate cyclase 9 (AC9), which is upregulated during neuronal differentiation, as a key ER-Ca2+-reducing regulator. Mechanistically, AC9-mediated production of cAMP is not essential for its ability to reduce ER Ca2+ content. Instead, AC9 inhibits store operated calcium entry (SOCE) by acting on Orai1, ultimately causing attenuation of ER Ca2+ level. More physiologically relevant, upregulation of AC9 in neurons is essential for reducing ER Ca2+ levels during Drosophila brain development. Collectively, this study lays a solid groundwork for further in-depth exploration of the regulatory mechanisms dictating ER Ca2+ homeostasis during neuronal differentiation and brain development.

cell biology↗

bioRxiv2023

Molecular understanding of anthocyanin biosynthesis activated by PAP1 in engineered red Artemisia annua cells and regulation of 2, 4-dichlorophenoxyacetic acid

Artemisia annua is an effective antimalarial medicinal crop. We have established anthocyanin-producing red cell cultures from this plant with the overexpression of Production of Anthocyanin Pigment 1 (PAP1) encoding a R2R3MYB transcription factor. To understand the molecular mechanism by which PAP1 activated the entire anthocyanin pathway, we mined the genomic sequences of A. annua and obtained eight promoters of the anthocyanin pathway genes. Sequence analysis identified four types of AC cis-elements from six promoters, the MYB response elements (MRE) bound by PAP1. In addition, six promoters were determined to have at least one G-Box cis-element. Eight promoters were cloned for activity analysis. Duel luciferase assays showed that PAP1 significantly enhanced the promoting activity of seven promoters, indicating that PAP1 turned on the biosynthesis of anthocyanins via the activation of these pathway gene expression. To understand how 2,4-dichlorophenoxyacetic acid (2,4-D), an auxin, regulates the PAP1-activated anthocyanin biosynthesis, five different concentrations (0, 0.05, 0.5, 2.5, and 5 M) were tested to characterize anthocyanin production and profiles. The resulting data showed that the concentrations tested decreased the fresh weight of callus growth, anthocyanin levels, and the production of anthocyanins per petri dish. HPLC-qTOF-MS/MS based profiling showed that these concentrations did not alter anthocyanin profiles. Real time RT-PCR was completed to characterize the expression PAP1 and four representative pathway genes. The results showed that the five concentrations reduced the expression levels of the constitutive PAP1 transgene and three pathway genes significantly and eliminated chalcone synthase gene in expression either significantly or slightly. These data indicate that the constitutive PAP1 expression depends on gradients added in the medium. Based on these findings, the regulation of 2,4-D is discussed for anthocyanin engineering in red cells of A. annua. ConclusionPromoters of eight anthocyanin pathway genes were cloned. Four types of AC cis-elements were identified from six promoters and G-Box elements were also determined from six promoters. PAP1 enhanced the activity of eight promoters. 2,4-D downregulated the expression of the constitutive PAP1 transgene leading to the decrease the biosynthesis of anthocyanins.

plant biology↗

bioRxiv2023

Measuring chondrocyte viability of articular cartilage based on label-free two-photon microscopy and deep learning image analysis

ObjectiveChondrocyte viability (CV) is an important indicator of articular cartilage health. Two-photon excitation autofluorescence (TPAF) and second harmonic generation (SHG) microscopy provide a label-free method for imaging chondrocytes. In this study, we propose an automated assessment of CV using deep learning cell segmentation and counting based on acquired TPAF/SHG images. DesignLabel-free TPAF/SHG images of cartilage samples from rats and porcine were acquired using both commercial and home-built two-photon microscopes, respectively. TPAF/SHG images were merged to form RGB color images with red, green, and blue channels assigned to TPAF (two channels) and SHG signals, respectively. To make the training datasets for the deep learning networks, individual chondrocyte areas on the RGB color images were manually circled and live or dead chondrocytes were validated by using Calcein-AM and Ethidium homodimer-1 dye labeling. We first built a chondrocyte viability network (MCV-Net) using the Mask R-CNN architecture, which could provide individual segmented cellular areas with live or dead status. Wiener deconvolution preprocessing was added before the input of MCV-Net to improve the accuracy of the CV analysis, forming the Wiener deconvolution CV network (wMCV-Net). ResultsTraining (300 images) and test (120 images) datasets were built for rats and porcine cartilage respectively. Wiener deconvolution could improve the Peak Signal-to-Noise Ratio (PSNR) for 30-40%. We demonstrated that both MCV-Net and wMCV-Net significantly improved the accuracy of the CV measurement. ConclusionA custom desktop TPAF/SHG microscope was used in collaboration with deep learning algorithm wMCV-Net based label-free method to assess the CV and get 95% accuracy with both rats and porcine samples.

bioengineering↗