This paper focuses on bearing fault diagnosis with limited training data. A major challenge in fault diagnosis is the infeasibility of obtaining sufficient training samples for every fault type under all working conditions. Recently deep learning based fault diagnosis methods have achieved promising results. However, most of these methods require large amount of training data. In this study, we propose a deep neural network based few-shot learning approach for rolling bearing fault diagnosis with limited data. Our model is based on the siamese neural network, which learns by exploiting sample pairs of the same or different categories. Experimental results over the standard Case Western Reserve University (CWRU) bearing fault diagnosis benchmark dataset showed that our few-shot learning approach is more effective in fault diagnosis with limited data availability. When tested over different noise environments with minimal amount of training data, the performance of our few-shot learning model surpasses the one of the baseline with reasonable noise level. When evaluated over test sets with new fault types or new working conditions, few-shot models work better than the baseline trained with all fault types. All our models and datasets in this study are open sourced and can be downloaded from https://mekhub.cn/as/fault_diagnosis_with_few-shot_learning/. INDEX TERMS Deep learning, few-shot learning, bearing fault diagnosis, limited data.

Background Aging is the strongest risk factor for Parkinson's disease (PD), which is a clinically heterogeneous movement disorder with highly variable age at onset. DNA methylation age (DNAm age) is an epigenetic clock that could reflect biological aging. Objectives The aim was to evaluate whether PD age at onset is associated with DNAm‐age acceleration (difference between DNAm age and chronological age). Methods We used the genome‐wide Infinium MethylationEPIC array to assess DNAm age in discovery (n = 96) and replication (n = 182) idiopathic PD cohorts and a unique longitudinal LRRK2 cohort (n = 220) at four time points over a 3‐year period, comprising 91 manifesting and 129 nonmanifesting G2019S carriers at baseline. Cox proportional hazard regression and multivariate linear regression were used to evaluate the relation between DNAm‐age acceleration and PD age at onset, which was highly variable in manifesting G2019S carriers (36–75 years) and both idiopathic PD cohorts (26–77 and 35–81 years). Results DNAm‐age acceleration remained steady over the 3‐year period in most G2019S carriers. It was strongly associated with age at onset in the LRRK2 cohort (P = 2.25 × 10−15) and discovery idiopathic PD cohort (P = 5.39 × 10−9), suggesting that every 5‐year increase in DNAm‐age acceleration is related to about a 6‐year earlier onset. This link was replicated in an independent idiopathic PD cohort (P = 1.91 × 10−10). In each cohort, the faster‐aging group has an increased hazard for an earlier onset (up to 255%). Conclusions This study is the first to demonstrate that DNAm‐age acceleration is related to PD age at onset, which could be considered in disease‐modifying clinical trials. Future studies should evaluate the stability of DNAm‐age acceleration over longer time periods, especially for phenoconverters from nonmanifesting to manifesting individuals. © 2022 International Parkinson and Movement Disorder Society.

DYRK1A, dual-specificity tyrosine phosphorylation-regulated kinase 1A, which is linked to mental retardation and microcephaly, is a member of the CMGC group of kinases. It has both cytoplasmic and nuclear functions, however, molecular mechanisms of how DYRK1A regulates gene expression is not well understood. Here, we identify two histone acetyltransferases, p300 and CBP, as interaction partners of DYRK1A through a proteomics study. We show that overexpression of DYKR1A causes hyperphosphorylation of p300 and CBP. Using genome-wide location (ChIP-sequencing) analysis of DYRK1A, we show that most of the DYRK1A peaks co-localize with p300 and CBP, at enhancers or near the transcription start sites (TSS). Modulation of DYRK1A, by shRNA mediated reduction or transfection mediated overexpression, leads to alteration of expression of downstream located genes. We show that the knockdown of DYRK1A results in a significant loss of H3K27acetylation at these enhancers, suggesting that DYRK1A modulates the activity of p300/CBP at these enhancers. We propose that DYRK1A functions in enhancer regulation by interacting with p300/CBP and modulating their activity. Overall, DYRK1A function in the regulation of enhancer activity provides a new mechanistic understanding of DYRK1A mediated regulation of gene expression, which may help in better understanding of the roles of DYRK1A in human pathologies.

Inflammatory bowel diseases (IBD), mainly including Crohn's disease and ulcerative colitis, are characterized by chronic inflammation of the gastrointestinal tract. Despite improvements in detection, drug treatment and surgery, the pathogenesis of IBD has not been clarified. A number of miRNAs have been found to be involved in the initiation, development and progression of IBD, and they may have the potential to be used as biomarkers and therapeutic targets. Here, we have summarized the recent advances about the roles of miRNAs in IBD and analyzed the contribution of miRNAs to general diagnosis, differential diagnosis and activity judgment of IBD. Furthermore, we have also elaborated the promising role of miRNAs in IBD-related cancer prevention and prognosis prediction.