Biology subjects

Xie, R.

Publications and source records attributed to Xie, R..

At least 19 recordsLinked to original sources

bioRxiv2026

Atlas-scale single-cell analysis beyond in-memory paradigm with scAtlasPy

Single-cell atlases are rapidly outgrowing the memory capacity of standard workstations, challenging the in-memory paradigm underlying mainstream computational ecosystems. Here, scAtlasPy decouples scale of atlas from memory capacity by leveraging the disk-resident computing. It enables full-resolution analysis of a 100-million-cell atlas with only 42.9 GB peak memory, whereas state-of-the-art platforms are limited at 3 million cells with 512 GB memory. scAtlasPy achieves 137,745 cells/s, 10.4x faster than scDataset with 82.6% lower memory usage for random minibatch retrieval. Its extensible architecture offers a flexible platform for diverse atlas-scale analytical tasks, facilitating the discovery of complex cellular heterogeneity and functions in massive cell atlases.

bioinformatics↗

bioRxiv2026

Presynaptic mitochondria calcium uniporter promotes auditory temporal processing during sustained high-rate activity

Mitochondrial calcium uniporter (MCU) uptakes calcium into mitochondria to maintain intracellular calcium homeostasis, malfunction of which has been implicated in altered neuronal signaling and disease. Its role in synaptic transmission remains understudied, especially in intact neural circuits. We investigated MCU function at the auditory nerve endbulb of Held synapse and postsynaptic bushy neurons in the cochlear nucleus, using age-matched control and MCU knockout (KO) mice of either sex. Whole-cell voltage-and current-clamp recordings were acquired from acute brain slices to examine synaptic transmission and postsynaptic responses. We found that basal synaptic properties at the endbulb of Held were unchanged in MCU KO mice, whereas synaptic transmission during sustained high-rate activity was significantly altered with a shift toward increased asynchronous release. Similarly, MCU deficiency did not change the intrinsic membrane properties of postsynaptic bushy neurons, but significantly reduced the temporal precision of auditory nerve evoked spikes trains at high rates. These results demonstrate that MCU is largely dispensable under low-rate activity, presumably because its activation requires relatively high calcium concentrations. In contrast, during sustained high-rate activity, MCU becomes an important regulator of synaptic function by reducing asynchronous neurotransmitter release under elevated intracellular calcium. Particularly in the auditory system, where neurons routinely fire at high rates, MCU promotes temporal processing and thereby plays a key role in supporting auditory function. It suggests that impaired MCU function under pathological conditions may be an important mechanism underlying central auditory processing deficits, and consequently contributes to hearing loss. SIGNIFICANT STATEMENTMCU uptakes calcium into mitochondria to regulate intracellular calcium, yet its contribution to synaptic transmission and neural processing remain understudied. Using a MCU KO mouse model, we investigated the role of MCU in synaptic transmission and postsynaptic responses at the endbulb of Held synapses and postsynaptic bushy neurons in the cochlear nucleus. The results showed that MCU plays little role in basal synaptic transmission, but significantly reduces asynchronous vesicle release during sustained high-rate activity, consequently improving temporal precision of the signal processing. These findings define the roles of MCU under different activity levels at the intact endulb/bushy connection, and suggest that impaired MCU function may be a key mechanism underlying central auditory processing deficits, and consequently contributes to hearing loss.

neuroscience↗

bioRxiv2026

Batch Effect Correction for Neuroimaging Data with Heterogeneous Spatial Correlations

Magnetic resonance imaging scans are effective tools for unveiling brain structures and understanding pathology for complex neurodevelopmental and aging processes. The spatial correlations among various brain regions reveal critical information and insights into the mechanisms of brain functions and their associations with cognitive abilities. Large-scale neuroimaging studies that acquire or aggregate imaging scans from multiple sites have become increasingly popular. Doing so enhances the diversity of study samples and robustness of study findings, and increases the statistical power of any analysis conducted for the biological hypotheses of interest. However, collecting images across different sites introduces non-biological variability attributed to differences in imaging protocol and configurations, known as batch effects. While there are methods to perform this batch effect correction, there are limited methods that directly account for the spatial patterns found in images of the brain. We develop Covariance-Aware Multivariate (CAM) ComBat that accounts for such spatial correlation across high-dimensional features, which could be heterogeneous across batches. We also propose a computationally efficient alternative of CAM-ComBat, Spatially-Informed Iterative Block (SIB) ComBat, that is scalable for very high dimension of features. We show that these methods outperform existing batch effect correction methods through simulation studies and an application to real neuroimaging data.

neuroscience↗

bioRxiv2026

Multiscale volume electron microscopy of the human liver maps vascular-cellular architecture, organelle dynamics and inter-organelle communication

The human liver depends on multiscale structural organization from vasculature to cells to organelles to perform its diverse metabolic functions. A unified three-dimensional view linking these hierarchical scales in intact human tissue would be useful for better understanding these levels and how they relate to each other. We present a high-resolution volume electron microscopy reconstruction of human periportal liver tissue (152 x 140 x 33 {micro}m3 at 8 nm pixel size) acquired by serial block-face scanning electron microscopy. Using a multiscale deep learning approach, we performed automated segmentation across the entire volume, enabling comprehensive annotation of vasculature, cells, and organelles. Quantitative analysis of bile duct architecture revealed coordinated scaling between lumen geometry and cholangiocyte number and size. Sinusoidal capillary branches exhibited distinct structural profiles with differential endothelial coverage. Analysis of 35,790 complete mitochondria identified substantial morphological heterogeneity, with elongated mitochondria displaying preferential endoplasmic reticulum (ER) contacts at narrowing sites, a pattern consistent with ER-mediated fission or fusion. This multiscale reconstruction establishes an ultrastructural reference for the healthy human liver and provides a quantitative framework for investigating hepatic physiology and disease.

cell biology↗

bioRxiv2026

An injectable soft implant for long-acting, reversible, ultra-stable release of therapeutics

Providing long-term (>6 months) zero-order drug release from easily administered formulations is a key challenge in improving patient adherence and facilitating access. Herein, we report the design and development of an injectable, biodegradable, long-acting polymeric microparticle-embedded hydrogel platform for prolonged, zero-order release of therapeutics. This "soft implant" is injectable for ease of administration and can be retrieved via a small incision, allowing for discontinuation of therapy if desired. Central to the platform are surface-eroding poly(orthoester) (POE) microparticles, which were molecularly tailored to tune zero-order drug release across a wide range of timeframes. We demonstrate the clinical potential of the "soft implant" using levonorgestrel, a contraceptive agent requiring sustained dosing. In vitro, we observed zero-order release for 300 days, projected for >12 months, with behavior consistent with surface erosion further supported through Raman chemical mapping. In vivo studies confirmed zero-order release for six months, projected to 12 months, from a subcutaneous injection in rats. We envision that our platform could transform therapies that require long-term, regular drug dosing, significantly improving compliance and therapy outcomes.

bioengineering↗

bioRxiv2026

Salmonella exploits metal-responsive UMPylation to enable intracellular biphasic growth and antibiotic tolerance

Salmonella develops profound antibiotic tolerance during in vivo infection, directly contributing to antimicrobial treatment failure. Although phenotypic heterogeneity in bacterial growth rates underlies this tolerance, the exact mechanisms remain elusive. Here, we reveal that Salmonella exhibits a biphasic growth pattern within macrophages, with initial antibiotic levels modulating lag phase duration. Crucially, this growth plasticity is orchestrated by metal ion homeostasis and a reversible PTM-mediated protein phase transitions. Mechanistically, under stress conditions, enhanced Mn2+ uptake activates the UMPylator YdiU, which modifies critical translational machinery components such as elongation factor TufA. These modifications induce protein aggregation, leading to translational arrest and growth stasis. Conversely, upon stress alleviation, Salmonella accumulates Mg2+ and effluxes Mn2+, then the de-UMPylase YdiV is activated, which resolves protein aggregates to enable rapid bacterial resuscitation and proliferation. This metal-responsive proteostatic switch represents an evolutionarily conserved strategy for bacterial stress adaptation, with broad implications for combating antimicrobial tolerance. SIGNIFICANCE STATEMENTAntibiotic tolerance, a key cause of treatment failure in chronic infections, is often linked to bacterial phenotypic heterogeneity, yet the underlying regulatory mechanisms remain largely understood. This study uncovers a fundamental survival strategy employed by intracellular pathogen Salmonella, where host-derived manganese acts as a critical signal to induce a reversible, dormant state. We demonstrate that stress-induced Mn2+ influx activates widespread protein UMPylation, which functions as a molecular glue to drive the aggregation of essential proteins, thereby halting bacterial growth and conferring antibiotic tolerance. Crucially, this state is reversible upon stress relief through Mn2+ efflux and subsequent de-UMPylation, enabling rapid bacterial resuscitation. As an evolutionarily conserved mechanism of adaptive tolerance, this metal-ion-responsive switch, mediated by reversible protein UMPylation, identifies a key target for developing therapeutics against persistent bacterial infections.

microbiology↗

bioRxiv2026

Lysosomal Aspartylglucosaminidase Couples Glycoprotein Catabolism to Cell-Surface GlycoRNA Production

GlycoRNAs, small non-coding RNAs covalently linked to N-glycans, reveal an unexpected intersection between RNA biology and glycosylation, with potential roles in immunity and cell-cell communication. The origins and biogenesis of their glycans remain unclear. Using CRISPR-Cas9 screening, metabolic glycan labeling and lysosome-targeted proteomics, we reconfirm canonical N-glycosylation enzymes (STT3A/B) and identify components of the mannose-6-phosphate (M6P) lysosomal sorting pathway (GNPTAB, M6PR) as essential contributors. Strikingly, lysosomal aspartylglucosaminidase (AGA) acts as a rate-limiting regulator, linking glycoprotein catabolism to glycoRNA formation. Loss of AGA collapses both abundance and diversity of RNA-linked N-glycans, while pulse-chase tracing shows glycoprotein-derived glycans are repurposed for RNA glycosylation. These findings position lysosomes as active participants in RNA modification and redefine glycoRNA biogenesis as a metabolically integrated process.

cell biology↗

bioRxiv2025

Rapid ecosystem collapse and biofilter formation following seabed methane leakage

Globally, vast amounts of methane are trapped within subseafloor gas hydrates. Recent evidence suggests that climate- and human-induced perturbations can destabilize gas hydrates and trigger methane release, yet the response of deep-sea ecosystems to these changes remains poorly understood. Here we show multi-year in situ monitoring of seabed ecosystem following methane leakage induced by hydrate exploration activities. Within just two years, benthic microbial and eukaryotic diversity in the affected areas declined significantly, while microbial and macrofaunal abundance increased. Integrated geochemical and omics analyses reveal the rapid successional shift of seabed ecosystem to a novel chemosynthetic system. Aerobic methanotrophs (Methyloprofundus) co-established with unexpectedly fast growing anaerobic methanotrophic archaea (ANME-3), accompanied by a rapid recruitment of opportunistic polychaetes that bioirrigated the seabed to >50 cm depths. The active animal-microbe interactions sustained exceptionally high rates of methane oxidation that utilized diverse electron acceptors. We demonstrate that methane hydrate destabilization can trigger rapid collapse of native seabed ecosystem while driving the formation of an effective methane biofilter consuming this potent greenhouse gas much faster than previously estimated. Understanding seabed ecosystem response and feedback is critical for predicting benthic methane cycling under ongoing global change and for developing sustainable strategies for methane hydrate resource management.

ecology↗

bioRxiv2025

Label-free biochemical imaging and timepoint analysis of neural organoids via deep learning-enhanced Raman microspectroscopy

Three-dimensional organoids have emerged as powerful models for studying human development, disease and drug response in vitro. Yet, their analysis remains constrained by standard imaging and characterisation techniques, which are invasive, require exogenous labelling and offer limited multiplexing. Here, we present a non-invasive, label-free imaging platform that integrates Raman microspectroscopy with deep learning-based hyperspectral unmixing for unsupervised, spatially resolved biochemical analysis of neural organoids. Our approach enables high-resolution mapping of cellular and subcellular structures in both cryosectioned and intact organoids, achieving improved imaging accuracy and robustness compared to conventional methods for hyperspectral analysis. Using our platform, we demonstrate volumetric imaging of a neural rosette within a neural organoid, and interrogate changes in biochemical composition during early developmental stages in intact neural organoids, revealing spatiotemporal variations in lipids, proteins and nucleic acids. This work establishes a versatile framework for high-content, label-free (bio)chemical phenotyping with broad applications in organoid research and beyond.

bioinformatics↗

bioRxiv2025

X-intNMF: A Cross- and Intra-Omics Regularized NMF Framework for Multi-Omics Integration

MotivationThe rapid accumulation of multi-omics data presents a valuable opportunity to advance our understanding of complex diseases and biological systems, driving the development of integrative computational methods. However, the complexity of biological processes--spanning multiple molecular layers and involving intricate regulatory interactions--requires models that can capture both intra- and cross-omics relationships. Most existing integration methods primarily focus on sample-level similarities or intra-omics feature interactions, often neglecting the interactions across different omics layers. This limitation can result in the loss of critical biological information and suboptimal performance. To address this gap, we propose X-intNMF, a network-regularized non-negative matrix factorization (NMF) model that explicitly incorporates both cross-omics and intra-omics feature interaction networks during multi-omics integration. By modeling these multi-layered relationships, X-intNMF enhances the representation of biological interactions and improves integration quality and prediction accuracy. ResultsFor evaluation, we applied X-intNMF to predict breast cancer phenotypes and classify clinical outcomes in lung and ovarian cancers using mRNA expression, microRNA expression, and DNA methylation data from TCGA. The results show that X-intNMF consistently outperforms state-of-the-art methods. Ablation studies confirm that incorporating both cross-omics and intra-omics interactions contributes significantly to the models improved performance. Additionally, survival analysis demonstrates that the integrated multi-omics representation offers strong prognostic value for both overall survival and disease-free status. These findings highlight X-intNMFs ability to effectively model multi-layered molecular interactions while maintaining interpretability, robustness, and scalability within the NMF framework. Availability and implementationSource code is available at: https://github.com/compbiolabucf/X-intNMF

bioinformatics↗

bioRxiv2025

Intratumoral microbiota-derived S1P sensitizes the combination therapy of Capecitabine and Nivolumab

Clinical responses of colorectal cancer (CRC) treatments vary considerably due to the heterogeneity of tumor micro-environment (TME), where intratumoral microbiota may reshape the unique inflammation imprints. However, its complex mechanistic underpinnings remain incompletely elucidated. Herein we sought to delineate the critical role of intratumoral microbiota in potentiating combination therapeutics against CRC. By comparing germ-free (GF) and specific pathogen-free (SPF) mouse models of 33 potential CRC treatments, we screened out the combination regimen of Capecitabine-Nivolumab significantly augmented by intratumoral microbiota in tumor regression and survival prolongation. The enrichment of Bacteroides fragilis induced by Capecitabine-Nivolumab was concomitant with elevated microbial sphingosine-1-phosphate (S1P), which further upregulated tumoral PD-L1 expression by enhancing histone deacetylation at the CD274 locus. Activation of microbial sphingosine kinase 2 (SphK2) ultimately led to an expansion of effector memory CD8+ T cells and reduction of exhausted T cell subsets within TME. To conclude, these findings advance our understanding of the intricate interplay among intratumoral microbiota, sphingolipid metabolites and immunochemotherapeutics. It highlights microbial sphingolipids as potential predictive biomarkers for strategies of targeting intratumoral microbiota in CRC management.

microbiology↗

bioRxiv2025

Activation of complement protein C3 depicts the dynamic tissue damage of the cochlea after noise trauma.

Hearing loss is characterized by dysfunction and pathological changes of the cochlea, including the loss of sensory hair cells and connected synapses. Growing evidence has revealed that the immune system plays critical roles in shaping cochlear responses to tissue injury and the clearance of damaged cells under pathological conditions such as noise-induced hearing loss (NIHL). However, the underlying mechanisms that regulate this process remain unclear. As a crucial part of the innate immune system, the complement system is widely known to respond to tissue damage and help remove cellular debris. Such functions are yet to be fully explored in the auditory system, especially in the cochlea. Among various complement factors, the activation of complement protein C3 is an essential functional hub of the complement cascade, resulting in the phagocytosis of dead or dying cells. In this study, we sought to investigate whether C3 exists in the cochlea and what roles it might play upon noise injury. To test this, we studied C3 activation using immunohistochemistry in the cochleae of various mouse models that were noise exposed to an octave band noise (8-16 kHz) for two hours at 112 dB SPL. Mice were sacrificed at various post exposure times to examine cochlear pathology. We found damage dependent C3 activity in the organ of Corti in mice after noise exposure (NE), especially in damaged outer hair cell (OHC) area. Specifically, C3 opsonized damaged cellular structures include degenerating OHCs, OHC debris, ribbon synapses, and efferent synaptic boutons, and the temporal dynamics of the opsonization closely resemble the timing of cochlear tissue damages post exposure. Unexpectedly, C3-deficient mice exhibited similar ABR thresholds and OHC counts at baseline as wild type mice, as well as comparable ABR threshold shifts and OHC loss after NE. However, we observed significantly reduced MOC terminal bouton degeneration in C3-deficient mice upon 112 dB SPL noise injury, indicating that C3 does modulate tissue damage and clearance, at least at the MOC terminals. Our results suggest that complement C3 is dynamically activated in the cochlea upon noise injury, which may play significant roles in damage recognition and further immune recruitment during the development of NIHL.

neuroscience↗

bioRxiv2025

Genome-Wide CRISPR Screening Identifies Cellular Factors Controlling Nonviral Genome Editing Efficiency

After administering genome editors, their efficiency is limited by a multi-step process involving cellular uptake, trafficking, and nuclear import of the vector and its payload. These processes vary widely across cell types and differ depending on the nature and structure of the vector, whether it is a lipid nanoparticle or a different synthetic material. We developed a novel genome-wide CRISPR screening strategy to better understand these limitations within human cells to identify genes modulating cellular uptake, payload delivery, and gene editing efficiency. Our screen interrogates the cellular processes controlling genome editing by Cas-based nuclease and base editing strategies in human cells. We designed a genome-wide screen targeting 19,114 genes in HEK293 cells, and we identified six genes whose knockout increased nonviral editing efficiency in human cells by up to five-fold. Further validation through arrayed knockouts of the top hits from our screen boosted the editing efficiency from 5% to 50% when Cas9 was delivered via lipid-based nanoparticles. By designing the guides to target the screen library cassette, we could accurately track the library sgRNA identity and the editing outcome on the same amplicon via short-read sequencing, enabling the identification of rare outcomes via computationally sorting edited from unedited cells within a heterogenous pool of >200M cells. In patient-derived human retinal pigment epithelium cells derived from pluripotent stem cells, BET1L, GJB2, and MS4A13 gene knockouts increased targeted genome editing by over five-fold. We anticipate that this high-throughput screening approach will facilitate the systematic engineering of novel nonviral genome editing delivery methods, where the identified novel gene hits can be further used to increase editing efficiency for other therapeutically relevant cell types.

genomics↗

bioRxiv2025

scGPT-spatial: Continual Pretraining of Single-Cell Foundation Model for Spatial Transcriptomics

Spatial transcriptomics has emerged as a pivotal technology for profiling gene expression of cells within their spatial context. The rapid growth of publicly available spatial data presents an opportunity to further our understanding of microenvironments that drive cell fate decisions and disease progression. However, existing foundation models, largely pretrained on single-cell RNA sequencing (scRNA-seq) data, fail to resolve the spatial relationships among samples or capture the unique distributions from various sequencing protocols. We introduce scGPT-spatial, a specialized foundation model for spatial transcriptomics continually pretrained on our previously published scGPT scRNA-seq foundation model. We also curate SpatialHuman30M, a comprehensive spatial transcriptomics dataset comprising of 30 million spatial transcriptomic profiles, encompassing both imaging- and sequencing-based protocols. To facilitate integration, scGPT-spatial introduces a novel MoE (Mixture of Experts) decoder that adaptively routes samples for protocol-aware decoding of gene expression profiles. Moreover, scGPT-spatial employs a spatially-aware sampling strategy and a novel neighborhood-based training objective to better capture spatial co-localization patterns among cell states within tissue. Empirical evaluations demonstrate that scGPT-spatial robustly integrates spatial data in mulit-slide and multi-modal settings, and effectively supports cell-type deconvolution and contextualized missing gene expression imputation, outperforming many existing methods. The scGPT-spatial codebase is publicly available at https://github.com/bowang-lab/scGPT-spatial.

bioinformatics↗

bioRxiv2025

RETINA: Reconstruction-based Pre-Trained Enhanced TransUNet for Electron Microscopy Segmentation on the CEM500K Dataset

Electron microscopy (EM) has revolutionized our understanding of cellular structures at the nanoscale. Accurate image segmentation is required for analyzing EM images. While manual segmentation is reliable, it is labor-intensive, incentivizing the development of automated segmentation methods. Although deep learning-based segmentation has shown promise, its accuracy to automatically segment cellular structures in EM data remains insufficient compared to expert manual results. Current approaches usually use either convolutional or transformer-based neural networks for image feature extraction. We developed the RETINA method, which combines pre-training on the large, unlabeled CEM500K EM image dataset with a hybrid neural-network model architecture that integrates both local (convolutional layer) and global (transformer layer) image processing to learn from manual image annotations. RETINA outperformed existing models on cellular structure segmentation on five public EM datasets. This improvement works toward automated cellular structure segmentation for the EM community. Data and code are available at Zenodo and GitHub.

bioinformatics↗

bioRxiv2025

The miR156/SPL9 interaction mode regulates the anthocyanin accumulation in potato

Anthocyanins are an essential class of flavonoids that represent a large group of plant secondary metabolites. microRNAs (miRNAs) can target TFs related to anthocyanin synthesis and inhibit their expression, thereby affecting the expression level of key downstream structural genes and ultimately regulating the synthesis and accumulation of anthocyanins in plants. Nevertheless, the regulatory function of miR156a in anthocyanin synthesis, especially in potatoes, has yet to be investigated. In this study, the miR156a and its target gene StSPL9 were screened and analyzed from small RNA sequencing, degradome sequencing and transient expression assay. The function of miR156a in anthocyanin synthesis of potato tuber was investigated. Overexpression of miR156a (OE-miR156a) in potato tubers promoted anthocyanin synthesis while concurrently reducing flavonoid synthesis compared with wild-type (WT) potatoes. The miR156a-silenced tubers (STTM-miR156a) achieved using short tandem target mimics (STTM) contained significantly lower anthocyanin content and increased flavonoid accumulation compared to WT and OE-miR156a. Notably, overexpression of StSPL9 in potato yielded identical results. The relative expression levels of the anthocyanin-related structural genes PAL, 4CL, CHS, CHI, F3H, DFR, and UFGT in STTM-miR156a transgenic potato tubers showed the opposite trend to that observed in OE-miR156a potato tubers, as demonstrated by RNA sequencing and quantitative real-time PCR (qRT-PCR). The integrated analysis of the transcriptome and metabolome revealed that the accumulation of flavonoids and flavonols in STTM-miR156a was significantly increased than in the wild type, whereas the contents of anthocyanins and phenolic acids were notably reduced. Further analysis showed that the expression of CHS, FG2, BEATH, and CYP75A were up-regulated in STTM-miR156a, and HCT was downregulated, which resulted in a reduced accumulation of methyl 4-caffeoylquinate, a phenolic acid metabolite. Therefore, it is demonstrated that miR156a indirectly modulates the expression of CHS, BEATH, CYP75A, HCT, and FG2 by targeting and degrading its target gene StSPL9 and other SPL genes, thus influencing anthocyanin accumulation in potato tuber. This study provides valuable insights into the complex regulatory network governing anthocyanin biosynthesis in Solanum species.

molecular biology↗

bioRxiv2024

Disentangling the CHAOS of intrinsic disorder in human proteins

Most proteins consist of both folded domains and Intrinsically Disordered Regions (IDRs). However, the widespread occurrence of intrinsic disorder in human proteins, along with its characteristics, is often overlooked by the broader communities of structural and molecular biologists. Building on the MobiDB database of intrinsic disorder in proteins, here we develop a comprehensive dataset (Comprehensive analysis of Human proteins And their disOrdered Segments - CHAOS). We implement internally consistent definitions of disordered regions, and annotate general characteristics such as cellular location, essentiality, post-translational modifications, and predicted pathogenicity. Further, we cross-reference to structure predictions from AlphaFold. We find that most human proteins contain at least one disordered region, predominantly located at the protein termini. IDRs are less hydrophobic, enriched in post-translational modifications, and mutations in IDRs are predicted to be less pathogenic than in non-IDRs. Additionally, we discovered that proteins residing in different cellular locations possess distinct disorder profiles. Finally, the predicted AlphaFold models of proteins in CHAOS suggest that disordered regions and proteins are often predicted to adopt secondary structure. Hereby we enhance the visibility and understanding of intrinsic disorder in human proteins. Key messagesO_LIFour out of five human proteins contain one or more intrinsically disordered regions (IDRs). C_LIO_LIHalf of the IDRs are located at protein termini, but three quarters of all human proteins contain a terminal IDR. C_LIO_LIThe amount and location of disordered regions differs throughout cellular compartments. C_LIO_LIOne in five missense mutations in IDRs are likely pathogenic. C_LIO_LIAlphaFold predicts secondary structure elements within intrinsically disordered regions and fully disordered proteins. C_LI

bioinformatics↗

bioRxiv2024

AlphaBridge: tools for the analysis of predicted macromolecular complexes

Artificial intelligence (AI)-powered protein structure prediction has transformed how scientists explore macromolecular function. AI-based prediction of macromolecular complexes is increasingly used for evaluating the likelihood of proteins forming complexes with other proteins, nucleic acids, lipids, sugars, or small-molecule ligands. Efficient tools are needed to evaluate these predicted models. We introduce an approach based on combining the confidence metrics of AlphaFold3 to enable clustering of sequence motifs participating in binary interactions and subsequently in 3D interfaces of complexes. Interaction interfaces within confidence limits are finally visualised in 2D using chord diagrams and network graphs. The analysis and visualisation are implemented in a web tool, which links them with interactive graphics and summary tables of predicted interfaces and intermolecular interactions, including confidence scores. Finally, we demonstrate real-life examples of how AlphaBridge is used for providing an efficient way to assess and validate predicted protein complexes and interfaces. The reproducible, objective and automated procedures we present provide a straightforward critical assessment of structure prediction of biomolecular complexes, that should be consulted before conducting more resource-intensive analyses.

bioinformatics↗