Here we describe the efficient generation of eGFP-transgenic rats using a lentiviral approach. Analysis of the founder generation demonstrated that 46% of the offspring had stably integrated the provirus into the genome and of those 92% expressed eGFP in all blood-derived leukocytes. In contrast to their offspring, all founder rats were mosaic with regard to eGFP-expression, suggesting delayed viral transduction after injection. The expression level of eGFP in the F1 generation is influenced by and segregates with the site of proviral integration. Interestingly, a single copy of the transgene is sufficient for reliable detection by flow cytometry, irrespective of the leukocyte subtype analyzed. Adoptive transfer of purified CD4(+) T-lymphocytes from transgenic rats and subsequent reisolation from various organs further demonstrated that expression of the lentiviral transgene is maintained in a foreign host and therefore allows for efficient tracking of transferred cells. Taken together, lentivirally generated eGFP-transgenic rats are a powerful tool for various applications in immunology and presumably also many other fields.

The achievement of the pathologic complete response (pCR) has been considered a metric for the success of neoadjuvant chemotherapy (NAC) and a powerful surrogate indicator of the risk of recurrence and long-term survival. This study aimed to develop a multimodal deep learning model that combined clinical information and pretreatment MR images for predicting pCR to NAC in patients with breast cancer. The retrospective study cohort consisted of 536 patients with invasive breast cancer who underwent pre-operative NAC. We developed a deep learning model to fuse high-dimensional MR image features and the clinical information for the pretreatment prediction of pCR to NAC in breast cancer. The proposed deep learning model trained on all datasets as clinical information, T1-weighted subtraction images, and T2-weighted images shows better performance with area under the curve (AUC) of 0.888 as compared to the model using only clinical information (AUC = 0.827, P < 0.05). Our results demonstrate that the multimodal fusion approach using deep learning with both clinical information and MR images achieve higher prediction performance compared to the deep learning model without the fusion approach. Deep learning could integrate pretreatment MR images with clinical information to improve pCR prediction performance.

The hepatitis B virus x protein (HBX) is expressed in HBVinfected liver cells and can interact with a wide range of cellular proteins. In order to understand such promiscuous behavior of HBX we expressed a truncated mini-HBX protein (named Tr-HBX) (residues 18-142) with 5 Cys Ser → mutations and characterized its structural features using circular dichroism (CD) spectropolarimetry, NMR spectroscopy as well as bioinformatics tools for predicting disorder in intrinsically unstructured proteins (IUPs). The secondary structural content of Tr-HBX from CD data suggests that Tr-HBX is only partially folded. The protein disorder prediction by IUPred reveals that the unstructured region encompasses its N-terminal ~30 residues of Tr-HBX. A two-dimensional 1 H-15 N HSQC NMR spectrum exhibits fewer number of resonances than expected, suggesting that Tr-HBX is a hybrid type IUP where its folded Cterminal half coexists with a disordered N-terminal region. Many IUPs are known to be capable of having promiscuous interactions with a multitude of target proteins. Therefore the intrinsically disordered nature of Tr-HBX revealed in this study provides a partial structural basis for the promiscuous structure-function behavior of HBX.

0000-0002-7464-9780 *These authors contributed equally to this work. Background/Aims: Human adenovirus type 55 (HAdV-55), an emerging epidemic strain, has caused several large outbreaks in the Korean military since 2014, and HAdV-associated acute respiratory illness (HAdV-ARI) has been continuously reported thereafter. Methods: To evaluate the epidemiologic characteristics of HAdV-ARI in the Korean military, we analyzed respiratory virus polymerase chain reaction (RV-PCR) results, pneumonia surveillance results, and severe HAdV cases from all 14 Korean military hospitals from January 2013 to May 2018 and compared these data with nationwide RV surveillance data for the civilian population. Results: A total of 14,630 RV-PCRs was performed at military hospitals. HAdV (45.4%) was the most frequently detected RV, followed by human rhinovirus (12.3%) and influenza virus (6.3%). The percentage of the military positive for HAdV was significantly greater than the percentage of civilians positive for HAdV throughout the study period, with a large outbreak occurring during the winter to spring of 2014 to 2015. The outbreak continued until the end of the study, and non-seasonal detections increased over time. The reported number of pneumonia patients also increased during the outbreak. Case fatality rate was 0.075% overall but 15.6% in patients with respiratory failure. The proportion of severe patients did not change significantly during the study period. Conclusions: A large HAdV outbreak is currently ongoing in the Korean military, with a trend away from seasonality, and HAdV-55 is likely the predominant strain. Persistent efforts to control the outbreak, HAdV type-specific surveillance, and vaccine development are required.

Background Electrocardiographic (ECG) monitors have been widely used for diagnosing cardiac arrhythmias for decades. However, accurate analysis of ECG signals is difficult and time-consuming work because large amounts of beats need to be inspected. In order to enhance ECG beat classification, machine learning and deep learning methods have been studied. However, existing studies have limitations in model rigidity, model complexity, and inference speed. Objective To classify ECG beats effectively and efficiently, we propose a baseline model with recurrent neural networks (RNNs). Furthermore, we also propose a lightweight model with fused RNN for speeding up the prediction time on central processing units (CPUs). Methods We used 48 ECGs from the MIT-BIH (Massachusetts Institute of Technology-Beth Israel Hospital) Arrhythmia Database, and 76 ECGs were collected with S-Patch devices developed by Samsung SDS. We developed both baseline and lightweight models on the MXNet framework. We trained both models on graphics processing units and measured both models’ inference times on CPUs. Results Our models achieved overall beat classification accuracies of 99.72% for the baseline model with RNN and 99.80% for the lightweight model with fused RNN. Moreover, our lightweight model reduced the inference time on CPUs without any loss of accuracy. The inference time for the lightweight model for 24-hour ECGs was 3 minutes, which is 5 times faster than the baseline model. Conclusions Both our baseline and lightweight models achieved cardiologist-level accuracies. Furthermore, our lightweight model is competitive on CPU-based wearable hardware.