Search

IntroductionProtein O-GlcNAcylation is a dynamic post-translational modification involved in major cellular processes and associated with many human diseases. Bioinformatic prediction of O-GlcNAc sites before experimental validation is a challenge task in O-GlcNAc research. Recent advancements in deep learning algorithms and the availability of O-GlcNAc proteomics data present an opportunity to improve O-GlcNAc site prediction.ObjectivesThis study aims to develop a deep learning-based tool to improve O-GlcNAcylation site prediction.MethodsWe construct an annotated unbalanced O-GlcNAcylation data set and propose a new deep learning framework, DeepO-GlcNAc, using Long Short-Term Memory (LSTM), Convolutional Neural Networks (CNN) combined with attention mechanism.ResultsThe ablation study confirms that the additional model components in DeepO-GlcNAc, such as attention mechanisms and LSTM, contribute positively to improving prediction performance. Our model demonstrates strong robustness across five cross-species datasets, excluding humans. We also compare our model with three external predictors using an independent dataset. Our results demonstrated that DeepO-GlcNAc outperforms the external predictors, achieving an accuracy of 92%, an average precision of 72%, a MCC of 0.60, and an AUC of 92% in ROC analysis. Moreover, we have implemented DeepO-GlcNAc as a web server to facilitate further investigation and usage by the scientific community.ConclusionOur work demonstrates the feasibility of utilizing deep learning for O-GlcNAc site prediction and provides a novel tool for O-GlcNAc investigation.

Cholesterol metabolism in the brain plays a crucial role in normal physiological function, and its aberrations are associated with cognitive dysfunction. The present study aimed to determine which cholesterol-related genes play a vital role in cognitive dysfunction and to dissect its underlying molecular mechanisms using a systems genetics approach in the BXD mice family. We first systematically analyzed the association of expression of 280 hippocampal genes related to cholesterol metabolism with cognition-related traits and identified lipoprotein lipase (Lpl) as a critical regulator. This was further confirmed by phenome-wide association studies that indicate Lpl associated with hippocampus volume residuals and anxiety-related traits. By performing expression quantitative trait locus mapping, we demonstrate that Lpl is strongly cis-regulated in the BXD hippocampus. We also identified ∼3,300 genes significantly (p < 0.05) correlated with the Lpl expression. Those genes are mainly involved in the regulation of neuron-related traits through the MAPK signaling pathway, axon guidance, synaptic vesicle cycle, and NF-kappa B signaling pathway. Furthermore, a protein–protein interaction network analysis identified several direct interactors of Lpl, including Rab3a, Akt1, Igf1, Crp, and Lrp1, which indicates that Lpl involves in the regulation of cognitive dysfunction through Rab3a-mediated synaptic vesicle cycle and Akt1/Igf1/Crp/Lrp1-mediated MAPK signaling pathway. Our findings demonstrate the importance of the Lpl, among the cholesterol-related genes, in regulating cognitive dysfunction and highlighting the potential signaling pathways, which may serve as novel therapeutic targets for the treatment of cognitive dysfunction.

Age-related hearing loss (ARHL) is associated with cognitive dysfunction; however, the detailed underlying mechanisms remain unclear. The aim of this study is to investigate the potential underlying mechanism with a system genetics approach. A transcriptome-wide association study was performed on aged (12–32 months old) BXD mice strains. The hippocampus gene expression was obtained from 56 BXD strains, and the hearing acuity was assessed from 54 BXD strains. Further correlation analysis identified a total of 1,435 hearing-related genes in the hippocampus (p < 0.05). Pathway analysis of these genes indicated that the impaired glutamatergic synapse pathway is involved in ARHL (p = 0.0038). Further gene co-expression analysis showed that the expression level of glutamine synthetase (Gls), which is significantly correlated with ARHL (n = 26, r = −0.46, p = 0.0193), is a crucial regulator in glutamatergic synapse pathway and associated with learning and memory behavior. In this study, we present the first systematic evaluation of hippocampus gene expression pattern associated with ARHL, learning, and memory behavior. Our results provide novel potential molecular mechanisms involved in ARHL and cognitive dysfunction association.

Objectives: To determine whether hypertension (HTN) affects cardiac structure and function in different types of hypertrophic cardiomyopathy (HCM). Design: Patients with obstructive HCM (n = 40), septal HCM (n = 88), and apical HCM (n = 42) were separated into hypertensive and non-hypertensive subgroups, and echocardiographic parameters at baseline and at follow-up were compared between the subgroups. Results: At follow-up, hypertensive obstructive HCM patients showed a decrease in end-diastolic volume (from 93.87 ± 26.08 mL to 79.06 ± 20.07 mL; p= 0.045) and in left ventricular end-diastolic diameter (from 45.00 ± 5.32 mm to 41.83 ± 4.58 mm; p =0.042). Non-hypertensive obstructive HCM patients showed a decrease in maximum aortic velocity (from 2.01 ± 0.53 m/s to 1.28 ± 0.25 m/s; p= 0.011) and in aortic maximum pressure gradient (from 17.22 ± 9.57 mm Hg to 6.79 ± 2.44 mm Hg; p= 0.03). Hypertensive apical HCM patients showed an increase in end-diastolic volume (from 95.28 ± 16.54 mL to 119.74 ± 25.19 mL; p= 0.016) and in left ventricular end-diastolic diameter (from 45.28 ± 3.36 mm to 50.20 ± 4.56 mm; p= 0.007). Conclusions: HTN can affect left ventricular capacity in obstructive HCM and apical HCM, causing a decrease in ventricular capacity in the former and increase in the latter; it has no significant effect on the size of the left ventricular cavity in septal HCM. HTN can lead to a poor therapeutic effect on aortic flow rate and pressure gradient in obstructive HCM patients.