Nat Commun. 2025 Aug 12;16(1):7436. doi: 10.1038/s41467-025-62579-z.

ABSTRACT

False positive claims of differentially expressed genes (DEGs) in scRNA-seq studies are of substantial concern. We found that DEGs from individual Parkinson's (PD), Huntington's (HD), and COVID-19 datasets had moderate predictive power for case-control status of other datasets, but DEGs from Alzheimer's (AD) and Schizophrenia (SCZ) datasets had poor predictive power. We developed a non-parametric meta-analysis method, SumRank, based on reproducibility of relative differential expression ranks across datasets, and found DEGs with improved predictive power. Specificity and sensitivity of these genes were substantially higher than those discovered by dataset merging and inverse variance weighted p-value aggregation methods. Up-regulated DEGs implicated chaperone-mediated protein processing in PD glia and lipid transport in AD and PD microglia, while down-regulated DEGs were in glutamatergic processes in AD astrocytes and excitatory neurons and synaptic functioning in HD FOXP2 neurons. Lastly, we evaluate factors influencing reproducibility of individual studies as a prospective guide for experimental design.

PMID:40796563 | DOI:10.1038/s41467-025-62579-z

J Proteome Res. 2025 Aug 12. doi: 10.1021/acs.jproteome.4c01060. Online ahead of print.

ABSTRACT

Bone tissue poses significant challenges for proteomic analysis due to its dense, mineral-rich matrix and predominance of collagen, overshadowing low-abundance proteins critical for understanding bone physiology during LC-MS/MS-based proteomic analysis. In this study, we present a rapid sequential two-step extraction protocol designed to enhance proteome coverage, reduce collagen interference without using collagenase, and ensure robust quantification while enabling simultaneous metabolome analysis. We systematically compared it with two previously reported methods, which attempt to reduce collagen content through enzymatic collagen digestion or by employing four sequential extractions. Performance was evaluated based on reproducible protein quantification, variance, collagen content, processing, and instrument time. Our protocol reproducibly quantified 4,518 proteins across a dynamic range of 4 orders of magnitude. It demonstrated only marginally inferior quantification performance compared to the four-step protocol while reducing extraction and measurement time by half. Further, it significantly outperformed the collagenase-based method, which quantified only 2,689 proteins. Incorporating a chloroform-methanol metabolite extraction only led to a minimal reduction in quantifiable proteins, making the protocol suitable for multiomics applications. In conclusion, this protocol facilitates comprehensive coverage of proteins after metabolite extraction, enabling comprehensive multiomics analyses and aiding in the assessment of bone diseases and therapeutic developments.

PMID:40796122 | DOI:10.1021/acs.jproteome.4c01060

Pharmacol Res. 2025 Aug 10:107905. doi: 10.1016/j.phrs.2025.107905. Online ahead of print.

ABSTRACT

Huntington's disease (HD) is a fatal neurodegenerative disorder caused by a CAG repeat expansion in the HTT gene. This leads to progressive loss of striatal neurons and motor-cognitive decline. While current gene-targeting approaches aiming at reducing somatic instability show promise - especially in case of early treatment - they cannot restore the already compromised neuronal circuitry at advanced disease stages. Thus, cell replacement therapy offers a regenerative strategy to rebuild damaged striatal circuits. Here, we report that human striatal progenitors (hSPs) derived from embryonic stem cells via a morphogen-guided protocol survive long-term when transplanted into a rodent model of HD and recapitulate key aspects of ventral telencephalic development. By employing single-nucleus RNAseq of the grafted cells, we resolved their transcriptional profile with unprecedented resolution. This has identified transcriptional signals of D1- and D2-type medium spiny neurons (MSN), Medial Ganglionic Eminence (MGE) and Caudal Ganglionic Eminence (CGE) -derived interneurons, and regionally specified astrocytes. Moreover, we demonstrate that grafted cells undergo further maturation 6 months post-transplantation, acquiring the expected regionally defined transcriptional identity. Immunohistochemistry confirmed stable graft composition over time and supported a neurogenic-to-gliogenic switch post-transplantation. Multiple complementary techniques including virus-based tracing and electrophysiology assays demonstrated anatomical and functional integration of the grafts. Notably, chemogenetic modulation of graft activity regulated striatal-dependent behaviors, further supporting effective graft integration into host basal ganglia circuits. Altogether, these results provide preclinical evidence that hSP-grafts can reconstruct striatal circuits and modulate functionally relevant behaviors. The ability to generate a scalable, molecularly defined progenitor population capable of in vivo functional integration supports the potential of hSPs for clinical application in HD and related basal ganglia disorders.

PMID:40796049 | DOI:10.1016/j.phrs.2025.107905

J R Soc Interface. 2025 Aug;22(229):20240877. doi: 10.1098/rsif.2024.0877. Epub 2025 Aug 13.

ABSTRACT

Physiological adaptation is a fundamental property of biological systems across all levels of organization, ensuring survival and proper function. Adaptation is typically formulated as an asymptotic property of the dose response (DR), defined as the level of a response variable with respect to an input parameter. In pharmacology, the input could be a drug concentration; in immunology, it might correspond to an antigen level. In contrast to the DR, this paper develops the concept of a transient, finite-time, cumulative dose response (cDR), which is obtained by integrating the response variable over a fixed time interval and viewing that integral-area under the curve-as a function of the input parameter. This study is motivated by experimental observations of cytokine accumulation under T-cell stimulation, which exhibit a non-monotonic cDR. It is known from the systems biology literature that only two types of network motifs, incoherent feedforward loops and negative integral feedback (IFB) mechanisms, can generate adaptation. Three paradigmatic such motifs-two types of incoherent loops and one integral feedback-have been the focus of much study. Surprisingly, it is shown here that these two incoherent feedforward loop motifs-despite their capacity for non-monotonic DR-always yield a monotonic cDR, and are therefore inconsistent with these experimental data. On the other hand, this work reveals that the IFB motif is indeed capable of producing a non-monotonic cDR, and is thus consistent with these data.

PMID:40795984 | DOI:10.1098/rsif.2024.0877

Cell Rep Methods. 2025 Aug 8:101140. doi: 10.1016/j.crmeth.2025.101140. Online ahead of print.

ABSTRACT

We identified a gene panel comprising 71 glycosyltransferases (GTs) that alter glycan patterns on cancer cells as they become more virulent. When these cancer-pattern GTs (CPGTs) were run through an algorithm trained on The Cancer Genome Atlas, they differentiated tumors from healthy tissue with 97% accuracy and clustered 27 cancers with 94% accuracy in external validation, revealing each variety's "biometric glycan ID." Using machine learning, we built four models for cancer classification, including two for detecting the molecular subtypes of breast cancer and glioma using even smaller CPGT sets. Our results reveal the power of using glyco-genes for diagnostics: Our breast cancer classifier was almost twice as effective in independent testing as the widely used prediction analysis of microarray 50 (PAM50) subtyping kit at differentiating between luminal A, luminal B, HER2-enriched, and basal-like breast cancers based on a comparable number of genes. Only four GT genes were needed to build a prognostic model for glioma survival.

PMID:40795869 | DOI:10.1016/j.crmeth.2025.101140

Curr Biol. 2025 Aug 6:S0960-9822(25)00952-2. doi: 10.1016/j.cub.2025.07.043. Online ahead of print.

ABSTRACT

The efficient recognition and removal of apoptotic cells and extracellular vesicles (EVs) by phagocytes is critical to prevent secondary necrosis and maintain tissue homeostasis. Such detection involves receptors and bridging molecules that recognize aminophospholipids-normally restricted to the inner leaflet of healthy cells-which become exposed on the surface of dead cells and the vesicles they produce.1,2,3,4,5 A majority of studies focus on phosphatidylserine (PS), for which there are well-established receptors that either bind to the lipid directly or indirectly via intermediary proteins.6,7,8 Phosphatidylethanolamine (PE) is even more prevalent than PS in the inner leaflet of mammalian cells9 and also becomes exposed by the action of scramblases during cell death,10,11 though little is known about the effects of PE once scrambled. Here, we report that PE can itself serve as a phagocytic ligand for macrophages by engaging CD300 family receptors. CD300a and CD300b specifically modulated the binding and uptake of PE particles, and this process involved immunoreceptor tyrosine-based activation motif (ITAM)-containing adaptors and spleen tyrosine kinase (Syk). For bacteria, which contain PE but largely lack PS in their membranes, we report that PE engagement enabled the binding and uptake of spheroplasts and bacterial extracellular vesicles (BEVs) that were unsheathed by the cell wall. The inflammatory responses of macrophages to PE particles containing lipopolysaccharide (LPS) were also curtailed by CD300a expression. Based on these observations, we posit that the direct recognition of PE facilitates mechanisms of clearance that stand to have a broad impact on the immune response.

PMID:40795848 | DOI:10.1016/j.cub.2025.07.043

Cell Rep Med. 2025 Aug 7:102283. doi: 10.1016/j.xcrm.2025.102283. Online ahead of print.

ABSTRACT

Melanoma is a deadly cancer with increasing incidence and mortality rates, and biomarkers for diagnosis are urgently needed. The impact of the microbiome, genetic factors, and immunologic markers on disease outcomes is described, but a comprehensive serum metabolome profiling is missing. The serum metabolome of patients with melanoma might be valuable to identify potential biomarkers. We present an untargeted metabolomics analysis in an exploratory cohort (87 patients with melanoma), an independent validation cohort (37 additional patients with melanoma featuring late-stage tumors), and 18 healthy control individuals, revealing striking differences. We identify and validate six serum metabolites that can predict the diagnosis of melanoma with an area under the curve (AUC) >0.9544 in advanced-stage melanoma. The AUC of our lead biomarker, muramic acid, is 0.964, 0.908, and 0.9936 in patients with stage I (n = 22), stage II (n = 67), and advanced melanoma (n = 86), respectively. In summary, we identify potentially very powerful diagnostic biomarkers for clinical practice.

PMID:40795845 | DOI:10.1016/j.xcrm.2025.102283

Semin Oncol. 2025 Aug 11;52(5):152396. doi: 10.1016/j.seminoncol.2025.152396. Online ahead of print.

ABSTRACT

Stomach adenocarcinoma (STAD) is one of the deadliest malignant tumors worldwide. Carbamoyl-phosphate synthetase 2 (CAD) expression is essential for categorizing and detecting STAD initiation and development. We explored the differential expression of genes (DEGs) affected by CAD overexpression and subsequently revealed the classification module of CAD-based scoring sets using weighted gene co-expression network analysis (WGCNA). Subsequently, enrichment analysis of biological functions and signaling pathways in clinically significant modules was conducted. We constructed a CAD-based clinical scoring model using univariate and multivariate Cox regression analyses. In addition, by using immune cell infiltration analysis, we investigated the interaction between CAD-based score and the immune microenvironment, identified upstream regulatory factors, including RNA binding proteins (RBPs), that affect the transcription of the STAD-related CAD-based score, and explored potential drug targets. We identified 4,977 abnormal regulatory genes related to CAD in STAD, among which the module genes most related to CAD were significantly enriched in cancer-related signaling pathways, such as VEGF, MAPK and TGF-beta signaling pathway. The CAD-based scores, T and N were identified as independent prognostic factors for STAD patients. We also found that under the influence of high expression of CAD, the infiltration level of most immune cells is lower, such as CD4 T cells and Tfh, and CAD has an inhibitory effect on the infiltration of certain immune cells. Notably, the potential drug targets PDHB and NDUFB6 are upstream regulatory factors in STAD. This study explored the role of highly expressed CAD-related genes in STAD and explored the tumorigenesis and progression of this disease. This research identified potential diagnostic and prognostic drug targets and provided new insights into the molecular mechanisms of STAD.

PMID:40795601 | DOI:10.1016/j.seminoncol.2025.152396

Microbiol Res. 2025 Aug 8;301:128308. doi: 10.1016/j.micres.2025.128308. Online ahead of print.

ABSTRACT

Mycovirus infections significantly impact fungal virulence and physiology, inducing either hypovirulence or hypervirulence. This study investigated the hypovirulent effects of Colletotrichum gloeosporioides RNA virus 1 (CgRV1-Ssa-44.1) infection on Colletotrichum gloeosporioides using multi-omics approaches. Transcriptomic analysis identified 261 differentially expressed genes (141 up-regulated, 120 down-regulated), while LC-MS/MS-based proteomic analyses revealed 2222 proteins, including 19 unique to virus-infected samples and 649 unique to virus-free samples. These results highlighted extensive gene and protein expression alterations, emphasizing profound impacts on the host cellular process. Changes in membrane-associated terms and cell wall-related processes suggested that the virus may exploit host structures to facilitate horizontal transfer. The disruption of carbohydrate metabolism and pathways, such as the non-sense mediated mRNA decay (NMD) system, reflected sophisticated viral strategies for suppressing host defenses and redirecting resources for its benefit. Notably, Upregulated genes, such as sorbose reductase and COMPASS complex component SWD2, pointed to adaptive response to stress and survival mechanisms during viral infection. Conversely, downregulated genes like elongation factor 3, survival factor 1, and zuotin, indicated viral manipulation of host cellular machinery to subvert normal processes. Real-time PCR validated these transcriptional changes, confirming the robustness of the findings. The study demonstrates a complex host-virus interplay, where fungal metabolic and adaptive pathways are intricately targeted and exploited. These findings underscore the dual nature of viral subversion strategies, balancing host suppression with survival adaptation. Future functional analyses of key pathways will provide insights into the molecular mechanisms underlying fungal-virus interactions and coevolution. This knowledge could guide the development of novel antifungal strategies applicable to similar host-pathogen systems.

PMID:40795492 | DOI:10.1016/j.micres.2025.128308

Sci Signal. 2025 Aug 12;18(899):eadr9131. doi: 10.1126/scisignal.adr9131. Epub 2025 Aug 12.

ABSTRACT

Increases in retroelement-derived double-stranded RNAs (dsRNAs) in various types of cancer cells facilitate the activation of antitumor immune responses. The long noncoding RNA EPIC1 interacts with the histone methyltransferase EZH2 and contributes to tumor immune evasion. Here, we found that EPIC1 in tumor cells suppressed cytoplasmic dsRNA accumulation, type I interferon (IFN) responses, and antitumor immunity. In various cancer cell lines, knockdown of EPIC1 stimulated the production of dsRNA from retroelements and an antiviral-like type I IFN response that activated immune cells. EPIC1 inhibited the expression of LINE, SINE, and LTR retroelements that were also repressed by EZH2, suggesting a potential role for the EPIC1-EZH2 interaction in regulating dsRNA production. In a humanized mouse model, in vivo delivery of EPIC1-targeting oligonucleotides enhanced dsRNA accumulation in breast cancer xenografts, reduced tumor growth, and increased the infiltration of T cells and inflammatory macrophages into tumors. Furthermore, EPIC1 knockdown improved the therapeutic efficacy of the immunotherapy drug pembrolizumab, a PD-1 inhibitor, in the humanized mouse model. Together, our findings establish EPIC1 as a key regulator of dsRNA-mediated type I IFN responses and highlight its potential as a therapeutic target to improve the efficacy of immunotherapy.

PMID:40794843 | DOI:10.1126/scisignal.adr9131

Ann Am Thorac Soc. 2025 Aug 12. doi: 10.1513/AnnalsATS.202501-028OC. Online ahead of print.

ABSTRACT

RATIONALE: Elexacaftor/tezacaftor/ivacaftor (ETI) has significantly improved lung function in people with cystic fibrosis (CF), prompting the need for outcome measures that can detect mild disease. In this new era of CFTR modulator therapy, more sensitive endpoints are required to evaluate the progression of early lung disease and to determine the efficacy of new CF therapies. Prior to the availability of highly effective therapies 129Xenon magnetic resonance imaging (Xe MRI) was shown to be more sensitive to regional ventilation changes compared to spirometry.

OBJECTIVES: To evaluate the longitudinal changes in pulmonary function and Xe-MRI outcomes after treatment with ETI in children and young people with CF.

METHODS: Lung function was assessed longitudinally at baseline 1, 6, and 12 months following ETI treatment initiation in children and young people with CF between the ages of 6 and 18 years at four study sites. Ventilation defect percentage (VDP), reader-defect percentage (RDP), Lung Clearance Index (LCI) and Forced Expiratory Volume in 1 second (FEV1) were reported.

MEASUREMENTS AND MAIN RESULTS: A total of 28 participants were enrolled; 25 completed at least baseline and one-month measurements. All four measures (RDP, VDP, LCI and FEV1) improved at one month after ETI initiation with a mean (standard deviation) absolute change of -1.2 (1.7) in LCI, 6.9 (12.3) in FEV1, -4.3 (4.8) in VDP and --7.8 (9.6) in RDP, respectively. Xe MRI outcomes (RDP and VDP) showed the largest relative treatment effects with mean relative improvements of 43% and 72%, respectively. One third of participants (8/25) had improvements in VDP and RDP but did not show improvements in FEV1.

CONCLUSIONS: Xe MRI captures sustained ventilation improvements following ETI initiation. Xe MRI metrics may provide a suitable endpoint for future interventional trials-particularly for people with CF with mild lung disease.

PMID:40795189 | DOI:10.1513/AnnalsATS.202501-028OC

Proc Natl Acad Sci U S A. 2025 Aug 19;122(33):e2506137122. doi: 10.1073/pnas.2506137122. Epub 2025 Aug 12.

ABSTRACT

Sweat parameters such as volume and chloride concentration may offer invaluable clinical insights for people with CF (PwCF). Pilocarpine-induced sweat collection for chloridometry measurement is the gold standard for a CF diagnosis, but this technique is cumbersome and not suitable for remote settings or repeat measurements. We have previously reported the utility of a skin-interfaced microfluidic device (CF Patch) in conjunction with a smartphone image processing platform that enables real-time measurement of sweating rates and sodium chloride loss in laboratory and remote settings. Here, we conducted clinical studies assessing the accuracy of the CF Patch compared to chloridometry when using pilocarpine to induce sweat. We also tested the feasibility and accuracy of exercise-induced sweat chloride measurements in PwCF and healthy volunteers (HV). In the laboratory, using either pilocarpine or exercise to induce sweat, the CF Patch demonstrated strong correlations with sweat chloride measured by pilocarpine-induced chloridometry. In remote settings, exercise-induced sweat chlorides measured using the CF patch were strongly correlated with in-laboratory exercise-induced CF patch sweat chlorides in HV but had a weaker correlation in PwCF. For PwCF on CFTR modulators, there was greater day-to-day variability in sweat chloride compared to HV, which highlights the limitations of assessing CFTR modulator efficacy and pharmacodynamics based on a single in-laboratory chloridometry measurement. Moreover, these findings demonstrate that the CF Patch is suitable as a remote management device capable of measuring serial sweat chloride concentrations and offers the potential of monitoring the efficacy of CF medication regimens but should not replace pilocarpine-based chloridometry for making a CF diagnosis.

PMID:40794828 | DOI:10.1073/pnas.2506137122

Sleep. 2025 Aug 8:zsaf226. doi: 10.1093/sleep/zsaf226. Online ahead of print.

ABSTRACT

STUDY OBJECTIVES: Atrial fibrillation (AF) and obstructive sleep apnea (OSA) are interrelated conditions that substantially increase the risk of cardiovascular complications. However, concurrent detection of these conditions remains a critical unmet need in clinical practice. Current home sleep apnea test (HSAT) devices often fail to detect arrhythmias essential for diagnosing OSA-associated AF due to limited ECG monitoring capabilities, and their integration with continuous positive airway pressure (CPAP) data for treatment optimization remains underutilized.

METHODS: This study introduces SHHDeepNet, an advanced deep learning-based framework designed for the detection of OSA in patients with AF, leveraging enhanced features extracted from single-lead electrocardiogram (ECG) signals. The ECG signals were preprocessed and refined using reconstruction independent component analysis (RICA), which isolates statistically independent features for improved data representation. These features were subsequently classified using the customized SHHDeepNet architecture. SHHDeepNet utilizes advanced signal processing and deep learning techniques to enhance ECG-based detection of AF-associated OSA.

RESULTS: The framework was validated using overnight ECG recordings from 101 subjects derived from the Sleep Heart Health Study Visit 1 (SHHS1) database, encompassing 36 prevalent AF (PAF) cases, 25 incident AF (IAF) cases, and 40 OSA cases. Detection performance was evaluated through binary classification (AF AH vs. AF non-AH) and multi-class classification (AF AH, AF non-AH, non-AF AH, and non-AF non-AH). During 5-fold cross-validation (5fold-CV), the framework achieved a binary classification accuracy of 98.22%, sensitivity of 96.8%, specificity of 99%, and an area under the curve (AUC) of 0.9981. For multi-class classification, 5fold-CV yielded 98.36% accuracy, 97.14% sensitivity, 98.77% specificity, and an AUC of 0.9975. Validation using leave-one-subject-out cross-validation (LOSO-CV) achieved a binary classification accuracy of 86.42%, sensitivity of 79.4%, specificity of 90.2%, and an AUC of 0.9372. For multi-class classification under LOSO-CV, the average accuracy, sensitivity, and F1-score were 86.7%, 72.6%, and 0.7224, respectively. External validation was performed on a cohort of 123 subjects from the Osteoporotic Fractures in Men (MrOS) database, comprising 68 cases of PAF and 55 cases of OSA. The proposed method achieved a multi-class classification accuracy of 88.51%, sensitivity of 73.50%, specificity of 91.34%, and an AUC of 0.9363.

CONCLUSIONS: These findings underscore the significance of simultaneous detection of AF and OSA, providing a more comprehensive evaluation of cardiovascular health. The proposed SHHDeepNet framework offers a promising tool to support clinical decision-making, enhance management strategies, and improve patient outcomes by mitigating the risks associated with these conditions.

PMID:40795334 | DOI:10.1093/sleep/zsaf226

J Gerontol A Biol Sci Med Sci. 2025 Aug 8:glaf173. doi: 10.1093/gerona/glaf173. Online ahead of print.

ABSTRACT

BACKGROUND: Chronological age is an important component of medical risk scores and decision-making. However, there is considerable variability in how individuals age. We recently published an open-source deep learning model to assess biological age from chest radiographs (CXR-Age), which predicts all-cause and cardiovascular mortality better than chronological age. Here, we compare CXR-Age to two established epigenetic aging clocks (First generation-Horvath Age; Second generation-DNAm PhenoAge) to test which is more strongly associated with cardiopulmonary disease and frailty.

METHODS: Our cohort consisted of 2,097 participants from the Project Baseline Health Study, a prospective cohort study of individuals from four US sites. We compared the association between the different aging clocks and measures of cardiopulmonary disease, frailty, and protein abundance collected at the participant's first annual visit using linear regression models adjusted for common confounders.

RESULTS: We found that CXR-Age was associated with coronary calcium, cardiovascular risk factors, worsening pulmonary function, increased frailty, and abundance in plasma of two proteins implicated in neuroinflammation and aging. Associations with DNAm PhenoAge were weaker for pulmonary function and all metrics in middle-age adults. We identified thirteen proteins that were associated with DNAm PhenoAge, one (CDH13) of which was also associated with CXR-Age. No associations were found with Horvath Age.

CONCLUSION: These results suggest that CXR-Age may serve as a better metric of cardiopulmonary aging than epigenetic aging clocks, especially in midlife adults.

PMID:40795299 | DOI:10.1093/gerona/glaf173

Sleep. 2025 Aug 7:zsaf227. doi: 10.1093/sleep/zsaf227. Online ahead of print.

ABSTRACT

STUDY OBJECTIVES: The intricate interplay between sleep and emotion has garnered increasing attention due to their profound impact on human health and well-being, including the development of interventions using emotion-regulating medications. While qualitative studies have illuminated their association, quantitative evidence remains limited.

METHODS: To address this gap, we leverage deep learning and emotion priors to explore the quantitative relationship between sleep and emotion using EEG signals. Our approach introduces novel emotion-based features into sleep stage classification, providing additional abstract information and corroborating the sleep-emotion link.

RESULTS: This method enables targeted interventions with emotion-regulating medications tailored to specific sleep stages. Furthermore, we investigate the quantitative influence of emotional combinations (emotional codings) on sleep stages, revealing distinct "emotional fingerprints" during sleep.

CONCLUSION: These findings support the development of corresponding drug combinations for sleep interventions. These findings lay the foundation for developing scientifically grounded and quantifiable approaches to sleep and emotion regulation, paving the way for advancements in understanding and addressing sleep and emotional disorders.

PMID:40795269 | DOI:10.1093/sleep/zsaf227

J Am Med Inform Assoc. 2025 Aug 6:ocaf118. doi: 10.1093/jamia/ocaf118. Online ahead of print.

ABSTRACT

OBJECTIVES: To improve prediction of chronic kidney disease (CKD) progression to end-stage renal disease (ESRD) using machine learning (ML) and deep learning (DL) models applied to integrated clinical and claims data with varying observation windows, supported by explainable artificial intelligence (AI) to enhance interpretability and reduce bias.

MATERIALS AND METHODS: We utilized data from 10 326 CKD patients, combining clinical and claims information from 2009 to 2018. After preprocessing, cohort identification, and feature engineering, we evaluated multiple statistical, ML and DL models using 5 distinct observation windows. Feature importance and SHapley Additive exPlanations (SHAP) analysis were employed to understand key predictors. Models were tested for robustness, clinical relevance, misclassification patterns, and bias.

RESULTS: Integrated data models outperformed single data source models, with long short-term memory achieving the highest area under the receiver operating characteristic curve (AUROC) (0.93) and F1 score (0.65). A 24-month observation window optimally balanced early detection and prediction accuracy. The 2021 estimated glomerular filtration rate (eGFR) equation improved prediction accuracy and reduced racial bias, particularly for African American patients.

DISCUSSION: Improved prediction accuracy, interpretability, and bias mitigation strategies have the potential to enhance CKD management, support targeted interventions, and reduce health-care disparities.

CONCLUSION: This study presents a robust framework for predicting ESRD outcomes, improving clinical decision-making through integrated multisourced data and advanced analytics. Future research will expand data integration and extend this framework to other chronic diseases.

PMID:40795063 | DOI:10.1093/jamia/ocaf118

Brief Bioinform. 2025 Jul 2;26(4):bbaf406. doi: 10.1093/bib/bbaf406.

ABSTRACT

This paper develops a deep neural network that accepts cell descriptors and molecules of multiple administered drugs and predicts the joint dose-response hypersurface of the combinatorial treatment. Since the dose-response hypersurface over several concentration dimensions fully characterizes the interaction dynamics of the administered drugs, the model is a computational tool that guides the discovery of synergistic treatments. The neural network is a biochemistry-informed universal approximator; it can estimate any shape of a dose-response hypersurface and has desirable invariances built into its architecture. The model excels at interpolating and extrapolating dose-response surfaces; its predictions align well with known mechanisms of action (MOA). It is the first model that can estimate joint dose-response hypersurfaces of arbitrarily many drugs, including untried combinations, in the presence of arbitrary, potentially nonlinear interactions between drugs. We release the model itself as well as a database of likely synergistic drug triplets. Our code is available at https://github.com/alonsocampana/PanThera/; the database of likely synergistic drug triplets at https://zenodo.org/records/14001717.

PMID:40794956 | DOI:10.1093/bib/bbaf406

Brief Bioinform. 2025 Jul 2;26(4):bbaf401. doi: 10.1093/bib/bbaf401.

ABSTRACT

Database searching remains the main approach for mass spectrometry-based proteomics, where protein identification fundamentally requires prior inclusion in the reference database. For aquatic algal species lacking annotated genomes, six-frame translation of species-specific transcriptomes has emerged as a prevalent method. However, this approach results in databases that encompass all potential translation products, substantially increasing the database size and search space. Here, we introduce BERT-based Protein Annotation (BPA), a deep learning strategy that combines a pretrained BERT model for contextual patterns, Pseudo Amino Acid Composition for physicochemical properties, and InterProScan for functional domain prediction, to optimize reference proteome construction. These features are integrated by using a Random Forest classifier to generate dynamic Sequence Reliability Scores, enabling adaptive filtering thresholds tailored to diverse experimental designs. Based on the validation across three distinct test species, this study demonstrates a robust performance of BPA with sustained high classification accuracy (AUC > 0.95). In the application to Karenia mikimotoi, BPA achieved 90% proteome compression while maintaining 40% identification coverage, effectively resolving the peptide ambiguity from redundant translations. This framework provides a scalable and efficient solution for constructing and optimizing reference libraries, facilitating proteomic research in aquatic algae and other genomically understudied species. Source code and executables are available at (https://github.com/huangruihua/BPA.git).

PMID:40794952 | DOI:10.1093/bib/bbaf401

Brief Bioinform. 2025 Jul 2;26(4):bbaf403. doi: 10.1093/bib/bbaf403.

ABSTRACT

The advancement of traditional Chinese medicine (TCM) faces challenges, due to the absence of a deep understanding of TCM mechanism at the perspective of modern biomedical practices. This results in how TCM selects herbs to treat diseases or symptoms prevailingly rely on clinicals' experience or TCM ancient books, at least in part lacking scientific basis. Herein, we present a novel deep learning-based approach, named Negative-Correlation-based TCM Architecture for Reversal (NeCTAR), to optimize the generation and combination of TCM formulas for guiding empiric therapy, by which we could, to some degree, narrow the gap between TCM and modern biomedical science. Our approach builds on a hypothesis that pathway alterations may serve as a proxy for the corresponding physiological changes induced by a certain disease, and 'inverse-fit' those alterations would provide a feasible therapeutic strategy to treat the disease. We leveraged ribonucleic acid sequencing (RNA-seq) data with Gene Set Enrichment Analysis to establish herb-pathway associations, integrating these insights into a multilayer perceptron model that incorporates top-k sparse projection and pathway reconstruction loss to predict the most therapeutically promising herbal components. NeCTAR demonstrated high concordance with experimental data across various disease models, including fatty liver disease, type 2 diabetes mellitus, and premature ovarian failure. Notably, NeCTAR could equally apply to single cell RNA-seq data. Overall, our study put forwards a novel interpretive framework underlying TCM mechanisms using modern biomedical foundation, by which we could prioritize herbal components based on existing TCM formulas treating diseases.

PMID:40794951 | DOI:10.1093/bib/bbaf403

Brief Bioinform. 2025 Jul 2;26(4):bbaf402. doi: 10.1093/bib/bbaf402.

ABSTRACT

Antigen presentation by major histocompatibility complex (MHC) molecules is a complex pathway essential for T cell-mediated immunity. The pathway involves unfolding and processing of the antigen protein structure, yet limited work has been made evaluating the potential influence of local protein structure on the prediction of antigen processing and presentation. Here, we investigated this by integrating local structural features-disorder score, relative surface accessibility, and the probabilities of α-helix, β-sheet, and coil-into an NNAlign-based framework for predicting MHC class I and HLA-DR antigen presentation. Large-scale eluted ligand datasets were used to train and validate our models, demonstrating that for MHC class I, the addition of structural features yielded marginal, nonsignificant improvements in performance. In contrast, for HLA-DR ligands, models incorporating positional structural information showed a significant yet limited performance boost. Post-hoc analysis revealed no clear amino acid enrichment patterns associated with structural propensities. Rather the HLA-specific gain in performance was found to be linked to the number of positive instances seen in training. Stratification by cellular localisation indicated that peptides from endolysosomal proteins benefited more from structural integration than those from cytosolic sources. Our comprehensive benchmark shows that incorporating local protein structural features improves epitope prediction for MHC class II ligands.

PMID:40794948 | DOI:10.1093/bib/bbaf402