Biology subjects

Pei, J.

Publications and source records attributed to Pei, J..

At least 19 recordsLinked to original sources

bioRxiv2026

AI-Based Comparison of UniProt annotations and Literature Across Human Kinases

Curated protein records and their supporting publications organize biological evidence at different levels of detail. Here, we used AI to independently answer a set of biological questions using either UniProt records or the full texts of the literature cited by those records, and then used AI-based evaluation to compare the resulting answers in terms of information coverage and semantic agreement. Fifty human proteins were selected from ten groups in a kinase sequence collection. Target-identity filtering retained 1,749 protein-publication records representing 1,655 distinct publications. For each protein, an AI language model generated one answer set from UniProt and another from the associated literature. A separate AI-based semantic evaluation compared all 2,500 pairwise combinations of the 50 UniProt-derived and 50 literature-derived answer sets, producing 87,500 question-level comparisons. Literature provided answers to 1,211 of 1,750 protein-question combinations, compared with 1,151 for UniProt. After correction for multiple testing across the 35 questions, literature coverage was significantly higher for autoinhibition and inhibitor type. Mean AI-assigned semantic agreement scores were higher for matched protein pairs (0.692) than for unmatched proteins within the same kinase group (0.384) and for proteins from different kinase groups (0.279). The matching literature-derived answer set ranked first for 49 of the 50 UniProt-derived queries, even after questions about gene symbol and kinase family were excluded. The sole exception was GUCY2C, whose UniProt-derived answer set showed slightly greater agreement with the NPR2 literature-derived answer set than with the GUCY2C literature-derived set. TYR proteins had the greatest overall answer coverage, whereas CK1 and RGC proteins had comparatively limited coverage in UniProt and the literature, respectively. Overall, AI analysis showed that UniProt records and their cited publications contain complementary yet strongly protein-specific biological information. Reporting gaps, question scope, and score aggregation influenced apparent agreement, underscoring that the framework evaluates consistency between AI-derived answers, not their biological accuracy.

bioinformatics↗

bioRxiv2026

Intranasal Fritillaria taipaiensis nanovesicles alleviate acute lung injury via the lung-brain axis

Acute lung injury (ALI) is a severe inflammatory syndrome frequently complicated by brain injury, which worsens outcomes. Here, we explored whether regulation of the lung-brain axis represents a therapeutic strategy for the management of ALI. We show that intranasal administration of Fritillaria taipaiensis-derived exosome-like nanovesicles (Ft-ELNs) ameliorates LPS-induced ALI and ALI-associated hypothalamic inflammation in mice. Ft-ELNs preferentially accumulate in the lungs, inhibiting TRPV1 in both the lungs and the hypothalamus to establish the lung-brain axis. In parallel, Ft-ELNs primarily regulate systemic humoral immunity by inhibiting the VEGFR1/PI3K/AKT/NF?B signaling pathway. In vitro experiments further confirmed that Ft-ELNs are efficiently internalized by alveolar epithelial MLE-12 cells, exerting protective effects by suppressing VEGFR1 expression. Mechanistically, we identified Ft-miR1 of Ft-ELNs as the key functional effector that directly targets the Flt1 3'UTR to inhibit VEGFR1 expression. These data suggest that Ft-ELNs attenuate lung-brain axis inflammation by targeting VEGFR1 and may hold potential as novel therapeutic agents for the treatment of severe ALI.

immunology↗

bioRxiv2026

Coacervates protect RNA against hydrolysis under wet-dry cycling conditions

RNA is crucial to all extant biology and thought to have played an essential role in lifes origins. However, hydrolysis of RNA under common environmental conditions such as heating, drying, and exposure to divalent cations presents challenges for its persistence and effectiveness in prebiotic contexts. RNAs with less stable secondary and tertiary structures are especially susceptible to degradation. In living cells, sequestration in biopolymer-rich membraneless organelles helps shield RNA from damage and enzymatic degradation. Here, we report that compartmentalization into polyelectrolyte-based complex coacervate droplets protects RNA from hydrolysis during wet-dry cycling at elevated temperatures in the presence of Mg2+. This protection can be understood in terms of the coacervate microenvironment under varied salt concentrations and how it changes during sample drying. After gaining insight into this prebiotically-plausible mechanism using an unstructured model RNA, we demonstrate coacervate-based protection of a ligase ribozyme, supporting the functional relevance of the protection.

biophysics↗

bioRxiv2026

Lineage recording reveals hijacked hepatic progenitor states as a common origin of HCC and ICC

HighlightsCERAMIC enables continuous and high-capacity lineage tracing of liver tumor initiation Fatty liver-associated hepatocytes acquire regenerative and premalignant cell states before malignant transformation Lineage reconstruction identifies Hep_Bi-zonal cells as the cellular origin of liver tumor initiation Transcriptional and regulatory programs distinguish tumor-fated hepatocytes from failed-to-transform lineages Peroxisomal metabolism is required for progenitor-state formation and liver tumor initiation Spatial remodeling identifies a macrophage niche associated with tumor-fated hepatocytes Dual ontogenies and functional specialization of lipid-associated macrophages shape the tumor-fated hepatocyte niche Fatty liver disease predisposes to primary liver cancer, yet the lineage routes and niche mechanisms that select rare tumor-fated hepatocytes remain unclear. Here we developed CERAMIC, a high-capacity CRISPR-Cas9 lineage recorder that co-recovers editing scars and transcriptomes from single cells, and applied it to an AKT/NRAS-driven model of MASLD-associated liver tumor initiation. Longitudinal lineage, single-cell and spatial analyses revealed a hierarchical trajectory in which bipotential bi-zonal hepatocytes (Hep_Bi-zonal), rather than pericentral-like hepatocytes (Hep_CVlike), generated regenerative and neoplastic hepatocyte progenitor states that progressed toward both hepatocellular carcinoma and intrahepatic cholangiocarcinoma lineages. Tumor-fated cells preferentially expanded along a remodeled midlobular-periportal axis and depended on ACOX1-mediated peroxisomal beta-oxidation to withstand lipotoxic and oxidative stress. Spatial and lineage analyses further identified a sequential lipid-associated macrophage niche, in which monocyte-derived LAMs engaged tumor-fated hepatocytes through an LGALS9-P4HB axis, and P4HB inhibition suppressed tumor expansion. These findings define liver tumor initiation as a lineage-restricted process licensed by peroxisomal metabolic adaptation and macrophage-derived niche signals.

cancer biology↗

bioRxiv2026

CellChem: Cellular transcriptional responses reshape molecular representation space for efficient and multi-scale drug discovery

Despite decades of progress in computational drug discovery, deep learning-based molecular representation models remain largely structure-centric, assuming that chemical similarity approximates functional similarity. However, drug effects in cells are shaped not only by chemical similarity but also by molecular interactions in the cellular context. To capture this complexity, we introduce CellChem, a cellular-chemical representation-learning framework that learns cell-guided molecular representations of small-molecule actions within cells. By incorporating large-scale cellular transcriptional profiles during pretraining, CellChem reshapes molecular representation space from structure-centric to a balanced integration of structure and function. The learned CellChem molecular representations exhibit biologically meaningful geometric organization, such that distances between molecules encode not only structural similarity but also similarity in the cellular responses they elicit, independent of downstream tasks. Using downstream derivative models such as cell-guided compound-protein interaction prediction and drug-induced transcriptional response profile generation, CellChem supports highly efficient, multi-scale drug discovery, achieving significantly better performance than traditional models.

molecular biology↗

bioRxiv2026

Domain classification of archaeal proteomes reveals conserved fold repertoire

Archaea represent one of the three domains of cellular life and yet account for fewer than 1% of experimentally determined protein structures, leaving the extent of their structural novelty unknown. Here we present a systematic domain-level classification of 124,075 proteins from 65 archaeal classes spanning 21 phyla and all major lineages, using both AFDB and newly predicted AlphaFold3 structures classified against the Evolutionary Classification of protein Domains (ECOD). We assigned 204,758 domains, of which 76.8% received high-confidence classifications, spanning 987 ECOD X-groups; 40% of known structural diversity within a single domain of life. Clustering by Foldseek recovered structural relationships for 63% of domains that are singletons by sequence comparison. To characterize the 21% of proteins lacking high-confidence classification, we applied successive filters for structure prediction confidence, protein length, and structural cluster context, reducing 8,452 domain-free proteins to a small number of well-folded structural orphans (less than 0.1% of the dataset). The unclassified fraction is dominated by sub-threshold matches to known folds (14% of all proteins) and low-confidence structure predictions (5%), not by novel structures. These results demonstrate that the protein fold repertoire at the single-domain level is broadly conserved across the deepest phylogenetic distances in cellular life, and that the gap between archaeal and well-characterized proteomes reflects classification sensitivity for divergent sequences rather than unexplored structural diversity.

bioinformatics↗

bioRxiv2026

Advanced Optical Microscopy Reveals Spatio-Temporal Dynamics of Cervix Remodeling during Gestation

Collagen remodeling in the uterine cervix is a vital process in pregnancy that allows for timely fetal delivery, yet its spatio-temporal details are still not fully understood. In this study, we measured collagen reorganization at different stages of murine gestation and at various cervical depths. We used polarization-resolved Second Harmonic Generation microscopy to specifically detect fibrillar collagen and assess its orientation with sub-micrometer resolution. We imaged large cervical areas using automated mosaicking and implemented an analysis pipeline that showed significant region-dependent changes in collagen quantity, porosity, and orientation disorder. Notably, we found that collagen disorganization begins in the lower cervix at gestation day 12 and extends throughout the entire cervix by day 15. Additionally, we demonstrated that the temporal dynamics of disorganization, without spatial sensitivity, can also be tracked using Mueller Matrix imaging, which is a clinically deployable method. These findings should improve understanding and diagnosis of gestation-related issues such as premature birth.

biophysics↗

bioRxiv2025

The inactivation of DHHC7 in mouse liver promotes diet-induced obesity through a hepatic Prg4-GPR146 axis

The liver secretes hepatokines that coordinate whole-body metabolism. Posttranslational lipidation of signaling proteins by DHHC palmitoyltransferases controls membrane targeting and signaling, yet the role of hepatic palmitoylation in systemic metabolic regulation is largely unexplored. Using an inducible, liver-specific DHHC7 knockout (D7LKO) mouse, we report that loss of DHHC7 in hepatocytes potentiates adenylyl cyclase-PKA-CREB signaling through reducing palmitoylation of inhibitory G protein subunit (Gi), leading to transcriptional elevation and secretion of proteoglycan 4 (Prg4). Elevated circulating Prg4 acts on adipocytes by binding GPR146 via Prg4s N-terminal SMB domain, suppressing adipocyte PKA substrate phosphorylation and hormone-sensitive lipase (HSL) Ser563 phosphorylation, thereby impairing lipolysis. Under high-fat diet (HFD) feeding, D7LKO mice and mice with adenoviral hepatic Prg4 overexpression develop pronounced obesity characterized by increasing brown, subcutaneous, and visceral fat mass and adipocyte hypertrophy; strikingly, these animals show little impairment in glucose tolerance or circulating triglycerides but display elevated plasma cholesterol. Conditioned medium containing Prg4 recapitulates suppression of adipocyte HSL phosphorylation in vitro, and SMB-deleted Prg4 fails to bind GPR146 or inhibit HSL phosphorylation. Our findings define a liver palmitoylation- hepatokine axis that controls adipose lipolysis and predisposes to diet-induced fat accumulation, establishing Prg4-GPR146 as a mechanistic link between hepatic signaling and adipose energy mobilization.

physiology↗

bioRxiv2025

Dynamic and Catalytic Multiphase Coacervates

Subcellular organization is critical for myriad cellular functions, including spatiotemporal regulation of biocatalysts. Liquid-liquid phase separation contributes to intracellular organization by providing microenvironments that sequester biomolecules and control their activities. Although catalytic reactions in multiphase coacervate systems are important for living cells, provide functionality to artificial systems, and may aid understanding of how metabolism first arose, they are not well understood. Here we report an artificial multiphase system in which catalytically active droplets containing three distinct coexisting coacervate phases undergo reorganization in response to pH, enabling control over chemical reactions. Catalysis is performed by an inner, polyhistidine-dense coacervate phase with inherent nonenzymatic activity. We show how phase composition, in addition to pH, affects catalytic activity, and how the addition of enzymes that alter local pH provides a means to regulate composition and catalysis. By exploiting competitive interplay between intramolecular self-association and intermolecular complexation among polymers within multiphase coacervate systems, it was possible to (re)organize and tune the reactivity of dynamic synthetic biological systems beyond what is possible for single-phase coacervates. Such approaches could prove valuable in designing artificial cells and related microreactor systems capable of catalyzing and controlling complex chemical and biochemical processes.

biochemistry↗

bioRxiv2025

UniCure: A Foundation Model for Predicting Personalized Cancer Therapy Response

Predicting drug efficacy across diverse patient contexts remains a major challenge in oncology, as models trained on cancer cell lines often fail to capture patient-specific biology. Emerging biological foundation models and patient-derived technologies offer a promising solution. Here, we present UniCure, the first pre-trained foundation model integrating both omics and chemical foundation models (UCE and Uni-mol) to predict transcriptomic responses to drugs across diverse cellular and tissue contexts, enabling personalized cancer therapy and drug prioritization at the individual level. Rather than encoder/decoder used in traditional models, UniCure utilizes parameter-efficient fine-tuning (PEFT) techniques for optimizing the training process, a novel FlexPert module for modeling flexible drug-cell interactions, and a Maximum Mean Discrepancy (MMD) loss for learning unpaired data. Trained on over 1.8 million perturbation RNA-seq profiles over 22,000 compounds, 166 cell types, and 24 tissues, UniCure achieves high accuracy in predicting both dose-dependent responses and drug combination effects, demonstrating strong generalization across bulk and single-cell transcriptomic data. In particular, fine-tuning on our patient-derived tumor-like clusters and real-world data of 800 profiles enables UniCure to generate individualized therapeutic predictions on patients tissue samples. In addition, UniCure enables patient stratification based on the predicted drug responses, providing a new way for subtyping patients. UniCures drug prioritization was validated across over 1000 patients from pan-cancer cohorts and supported by experiments of candidate therapeutics. By enabling the potential to screen millions of compounds per patient at scale, UniCure represents a biologically grounded tool that could advance personalized precision oncology and accelerate drug discovery.

bioinformatics↗

bioRxiv2025

Predicting Spatial Transcriptomics from H&E Image by Pretrained Contrastive Alignment Learning

The intricate molecular landscape within tissues holds crucial information about cellular behavior and disease progression, yet capturing this complexity at a spatial level remains challenging. While Spatial transcriptomics (ST) offers valuable insights into gene expression patterns within their native tissue context, its widespread adoption is hindered by high costs and limited gene detection capabilities. Here we introduce CarHE (Contrastive Alignment of gene expRession for hematoxylin and eosin image), a method that overcomes these limitations by accurately predicting high-dimensional ST data (over 10,000 genes) solely from readily available H&E (Hematoxylin and Eosin) stained images. This novel pre-trained architecture employs contrastive learning through two mechanisms: cell-type-based transcriptomics information transfer and image-based histology information transfer. These mechanisms precisely align image features with spatial single-cell gene expressions, achieving prediction accuracies exceeding 0.7 (up to 1.7 folds compared to second best) across diverse tissue types and species. CarHEs superior performance extends to identifying subtle pathological features such as tertiary lymphoid structures in various cancers, including breast cancer, lung cancer, melanoma and ccRCC (clear cell renal cell carcinoma), and reconstructing 3D spatial transcriptomics from images alone, offering a cost-effective and robust alternative for large-scale spatial transcriptomics. We further validated CarHEs effectiveness by predicting DFS (Disease-Free Survival) from >1,600 lung cancer patients HE images, achieving a significantly higher AUC (Area Under the Receiver Operating Characteristic Curve) of 0.73 compared to state-of-the-art alternatives (0.58-0.64).

bioinformatics↗

bioRxiv2025

Comparative assessment of co-folding methods for molecular glue ternary structure prediction

Molecular glues (MGs) represent an emerging therapeutic paradigm capable of inducing or stabilizing protein-protein interactions, with broad applications in creating neomorphic interactomes and targeted protein degradation. However, current discovery efforts remain largely confined to experimental screening, while in silico rational design of MGs persists as a formidable challenge. A critical step toward rational design lies in accurate ternary complex modeling. Given the scarcity of such complexes in the Protein Data Bank for training specialized predictive models, we tested the ability of recently developed co-folding models, including AlphaFold 3, Boltz-1, Chai-1, Protenix, and RoseTTAFold All-Atom in building the complex models. We systematically curated a dataset, named MG-PDB, with 221 non-covalent MG-engaged ternary complexes. MGBench were further introduced as a comprehensive benchmark set, which comprises 88 ternary structures excluded from co-folding models training data through rigorous time-based partitioning. Our benchmark results demonstrated that AlphaFold3 achieved the best overall performance among co-folding methods, in terms of both protein-protein interaction interface prediction (50.6% success rate) and MG-protein interaction recovery (32.9% success rate). However, our homology study showed that most of their successful predictions actually stemmed from memorization. Further analysis revealed three phenomena of current co-folding methods for MG ternary structure prediction. Firstly, these methods struggle to accurately model large interaction interfaces. Secondly, their predictive accuracy is notably reduced for domain-domain complexes compared to domain-motif interactions. Lastly, they face specific challenges in modeling MG degrader complexes with sufficient accuracy. We showcased they relied on the existing interaction patterns, and highlighted the need for further improved in novel E3 ligase systems. These findings reveal fundamental gaps in existing methods to learn atomic-level interaction rules for MG-engaged ternary complex modeling. MG-PDB and MGBench provide critical resources and mechanistic insights to advance computational MG discovery. Table of Contents Graphic O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=127 SRC="FIGDIR/small/655997v1_ufig1.gif" ALT="Figure 1"> View larger version (25K): org.highwire.dtl.DTLVardef@172b19borg.highwire.dtl.DTLVardef@ea4844org.highwire.dtl.DTLVardef@12ea246org.highwire.dtl.DTLVardef@9fb4bc_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2025

AMPK- and cGAS-Mediated Dual Sensing of Glucose and dsDNA Synergizes to Enhance Antitumor Immunity

Intratumoral glucose availability is linked considerably to tumor growth, immune evasion, and metastasis. However, the precise molecular mechanisms by which glucose levels govern cellular and immune responses against tumors remain incompletely understood. Here, we explored the context of tumors of intersections between AMP-activated protein kinase (AMPK) and cGAS-STING signaling, a key pathway initiating antitumor immunity. We unexpectedly found that AMPK activation robustly enhances cGAS-STING signaling within cancer and immune cells, which potently impacts melanoma and colorectal cell senescence, organoid apoptosis, and the infiltration of CD4+ and CD8+ T lymphocytes. Using classic or newly developed AMPK agonist Aldometanib, TBK1S511E/S511E knock-in (KI) mice, glucose metabolic modulator, as well as AMPK1/2 dKO melanoma, we demonstrated that dual sensing of intratumoral glucose and dsDNA integrated by the AMPK-TBK1 cascade is essential for initiating antitumor immunity and suppressing cancer progressions. Notably, novel AMPK agonists or intervening in glucose glycolysis by Lonidamine synergized effectively with STING agonists, substantially inhibiting melanoma growth. Therefore, these findings unravel a concise mechanism integrating glucose metabolism and innate DNA sensing into cancer cell fate and propose an effective therapeutic strategy to enhance antitumor immunity. HIGHLIGHTSO_LIDual and synergistic sensing of glucose and dsDNA facilitates antitumor immunity in melanoma and colorectal cancer; C_LIO_LIAMPK substitutes intratumoral cGAS-STING-IRF3 signaling through TBK1 phosphorylation at S511; C_LIO_LIGlucose deficiency induces cancer cell senescence, apoptosis, and T lymphocyte infiltration via the AMPK-TBK1 cascade; C_LIO_LINewly developed AMPK agonists synergize with STING agonists to retrieve striking tumor suppression. C_LI

cancer biology↗

bioRxiv2024

LRRK2 G2019S mutation incites increased cell-intrinsic neutrophil effector functions and intestinal inflammation in a model of infectious colitis

Parkinsons Disease (PD) is a progressive, neurodegenerative disorder characterised by motor and non-motor symptoms. Emerging evidence suggests a link between PD and gastrointestinal dysfunction. Constipation is frequently observed years prior to development of motor dysfunction in PD, and people with inflammatory bowel disease (IBD) are more likely to develop PD. Mutations in the leucine-rich repeat kinase 2 gene (LRRK2) account for approximately 1% of all PD cases and are associated with increased risk for IBD. Among them, LRRK2 Gly2019Ser (G2019S), located within the kinase domain, is the most common PD-associated mutation and increases kinase activity. It is unknown how LRRK2 mutation affects susceptibility to intestinal inflammation or pathogenesis of PD. Using single cell RNA sequencing (scRNAseq), we demonstrate that LRRK2 G2019S mutation promotes a dysregulated gene profile, especially within neutrophil, monocyte and {gamma}{delta} T cell populations, following Citrobacter rodentium infection in mice. Transcriptionally, LRRK2 G2019S neutrophils have a greater pro- inflammatory type I and II IFN response compared to those of WT mice. This is accompanied by an increase in neutrophil numbers in the lamina propria in LRRK2 G2019S mice. We also uncover cell-intrinsic functional defects in LRRK2 G2019S neutrophils, including increased chemotaxis, degranulation and neutrophil extracellular traps (NETosis) formation. Increased neutrophil infiltration is associated with an upregulation in Th17 immune responses, which may together contribute to the observed increase in colon pathology during infection. These findings increase our understanding of the role of PD-associated genes in immune cells and their contribution to immune dysregulation. Understanding the early perturbations driven by the LRRK2 G2019S mutation in gastrointestinal pathology may facilitate the development of biomarkers for early diagnosis and intervention in PD.

immunology↗

bioRxiv2024

Computing the Human Interactome

Protein-protein interactions (PPI) are essential for biological function. Recent advances in coevolutionary analysis and Deep Learning (DL) based protein structure prediction have enabled comprehensive PPI identification in bacterial and yeast proteomes, but these approaches have limited success to date for the more complex human proteome. Here, we overcome this challenge by 1) enhancing the coevolutionary signals with 7-fold deeper multiple sequence alignments harvested from 30 petabytes of unassembled genomic data, and 2) developing a new DL network trained on augmented datasets of domain-domain interactions from 200 million predicted protein structures. These advancements allow us to systematically screen through 200 million human protein pairs and predict 18,316 PPIs with an expected precision of 90%, among which 5,578 are novel predictions. 3D models of these predicted PPIs nearly triple the number of human PPIs with accurate structural information, providing numerous insights into protein function and mechanisms of human diseases.

biochemistry↗

bioRxiv2024

A rewiring of the earliest immune events leading to T-cell mediated disease following intestinal infection in a PINK1KO mouse model of Parkinsons disease

Parkinsons disease (PD) is characterized by a protracted period of non-motor symptoms, including gastrointestinal (GI) dysfunction, which can precede the development of the cardinal motor deficits by decades. This long prodrome of disease is highly suggestive of immune cell involvement in the initiation of disease, but currently the field lacks robust model systems to study such mechanisms. It has been hypothesized that pathology may be first initiated in the periphery due to environmental triggers, such as pathogens that enter the GI tract. We further speculate that the impact of such pathogens on the immune system could be exacerbated in genetically predisposed individuals. Our group has developed a GI-targeted pathogen-induced PD mouse model system in PINK1 KO mice with Gram-negative bacterial infections and found that T cells are a major player in driving PD-like motor symptoms at late stages following infection. Herein, we now map the initiating immune events at the site of infection at the earliest stages with the goal of shedding light on the earliest mechanisms triggering immune-mediated pathological processes relevant to PD. Using unbiased single cell sequencing, we demonstrate that myeloid cells are the earliest dysregulated immune cell type in PINK1 KO infected mice at 1-week post-infection, followed by a dysregulated T cell response shortly after, at 2 weeks post-infection. We find that these myeloid cells have an enhanced proinflammatory profile, are more mature, and develop enhanced capacity for antigen presentation. Using unbiased prediction analysis, our data suggest that cytotoxic T cells and myeloid cells are particularly poised for interacting with each other, and we identify possible direct cell-cell interaction pathways that might be implicated. Taken together, deciphering the earliest immune mechanisms in the periphery underpinning PD autoimmunity will be instrumental in the development of effective therapeutic targeting strategies before irrevocable neuronal damage ensues.

immunology↗

bioRxiv2024

Identification of Candidate Diagnostic gene CDC25C for Trametinib-induced Cardiotoxicity by Integrating Bioinformatics and Machine Learning

Trametinib is a MEK inhibitor that has been shown to have considerable efficacy in retarding melanoma progression. However, the exact mechanism of cardiotoxicity induced by trametinib remains unclear. Our study was designed to investigate the underlying mechanisms by which trametinib-induced cardiotoxicity (TIC) might exist, offering novel perspectives and guidance for potential prediction, diagnosis, and treatment of TIC. The GSE217421 dataset indicated 574 up-regulated and 705 down-regulated DEGs. According to the KEGG analysis, these genes were implicated in several pathways and functions, including Cell Cycle, Axon Guidance, Cellular Senescence, and Dilated Cardiomyopathy. The GO analysis suggested their association with Mitotic Cell Cycle, Microtubule Cytoskeleton and Adenyl Nucleotide Binding. The hub genes (CDC25C) for TIC were screened through Multiple machine learning algorithms. Next, The expression level of CDC25C was verified using the GSE217423 validation set. Nomogram model based on CDC25C demonstrated excellent diagnostic capability according to three different evaluation measures. To further explore the regulatory mechanism of CDC25C in TIC, we constructed a multi-regulatory network using miRNAs-lncRNAs-TFs-CDC25C and conducted an immunoinfiltration analysis. Our study suggests that CDC25C may be a candidate diagnostic gene and a potential therapeutic target for the early occurrence and development of TIC. This provides new ideas for the prediction, diagnosis and treatment strategies for TIC.

cancer biology↗

bioRxiv2024

Essential and virulence-related protein interactions of pathogens revealed through deep learning

Identification of bacterial protein-protein interactions and predicting the structures of the complexes could aid in the understanding of pathogenicity mechanisms and developing treatments for infectious diseases. Here, we developed a deep learning-based pipeline that leverages residue-residue coevolution and protein structure prediction to systematically identify and structurally characterize protein-protein interactions at the proteome-wide scale. Using this pipeline, we searched through 78 million pairs of proteins across 19 human bacterial pathogens and identified 1923 confidently predicted complexes involving essential genes and 256 involving virulence factors. Many of these complexes were not previously known; we experimentally tested 12 such predictions, and half of them were validated. The predicted interactions span core metabolic and virulence pathways ranging from post-transcriptional modification to acid neutralization to outer membrane machinery and should contribute to our understanding of the biology of these important pathogens and the design of drugs to combat them.

biochemistry↗