A Study on the Architecture of Artificial Neural Network Considering Injection-Molding Process Steps.

In this study, an artificial neural network (ANN) was established to predict product properties (mass, diameter, height) using six process conditions of the injection-molding process (melt temperature, mold temperature, injection speed, packing pressure, packing time, and cooling time) as input parameters. The injection-molding process consists of continuous sequential stages, including the injection stage, packing stage, and cooling stage. However, the related research tends to have an insufficient incorporation of structural characteristics based on these basic process stages. Therefore, in order to incorporate these process stages and characteristics into the ANN, a process-based multi-task learning technique was applied to the connection between the input parameters and the front-end of the hidden layer. This resulted in the construction of two network structures, and their performance was evaluated by comparing them with the typical network structure. The results showed that a multi-task learning architecture that incorporated process-level specific structures in the connections between the input parameters and the front end of the hidden layer yielded relatively better root mean square errors (RMSEs) values than a conventional neural network architecture, by as much as two orders of magnitude. Based on these results, this study has provided guidance for the construction of artificial neural networks for injection-molding processes that incorporates process-stage specific features and structures in the architecture.

Effects of Input Parameter Range on the Accuracy of Artificial Neural Network Prediction for the Injection Molding Process.

Artificial neural network (ANN) is a representative technique for identifying relationships that contain complex nonlinearities. However, few studies have analyzed the ANN's ability to represent nonlinear or linear relationships between input and output parameters in injection molding. The melt temperature, mold temperature, injection speed, packing pressure, packing time, and cooling time were chosen as input parameters, and the mass, diameter, and height of the injection molded product as output parameters to construct an ANN model and its prediction performance was compared with those of linear regression and second-order polynomial regression. Following the preliminary experiment results, the learning data sets were divided into two groups, i.e., one showed linear relation between the mass of the final product and the range of packing time (linear relation group), and the other showed clear nonlinear relation (nonlinear relation group). The predicted results of ANN were relatively better than those of linear regression and second-order polynomial for both linear and nonlinear relation groups in our specific data sets of the present study.

An Experimental Study on Color Shift of Injection-Molded Mobile LGP Depending on Surface Micropattern.

In the display industry, the LCD backlight unit (BLU) module is variously used in mobile phones, notebook computers, monitors, and TVs. The light guide plate (LGP), which is one of the core parts of a BLU, is getting bigger and thinner consistently. Conventional injection methods and injection processes like injection compression molding (ICM) are becoming more complex and harsher with high-speed injection at high mold and melt temperatures. These approaches lead to a change in physical properties and a decrease in optical properties such as yellowing and color shift in injection-molded parts. In the present study, an injection molding experiment was conducted to understand the effect of surface patterns and major injection process conditions like mold and melt temperatures on the color shift in injection-molded mobile LGP. Optical properties obtained by the direct and total transmittance and CIE xy chromaticity diagram for injection-molded mobile LGP were measured with a UV-visible spectrophotometer. From the measurement of patternless LGP, it was found that total or direct transmittance was not affected by molding process variables. It was also found that yellow shift, ΔE(xy), occurred as much as 0.00111 ± 0.00014, and a color shift angle, Θ(xy), of 43.05 ± 2.44° was recorded in CIE coordinates for all nine experimental cases. From the measurement of total transmittance of patterned LGP, ΔE(xy) and Θ(xx) were found to be almost the same as those of patternless LGP for the locations of low and medium density of the pattern for the LGP, T1 and T2. The measured data of direct transmittance of patterned LGP showed that additional yellow shift due to scattering caused by surface micropattern. Interestingly, Θ(xy) of patterned data remained 43.05 ± 2.44°, which was almost the same as that found in the case of patternless LGP.

Loss of Life Expectancy by 10 Years or More From Elevated Aspartate Aminotransferase: Finding Aspartate Aminotransferase a Better Mortality Predictor for All-Cause and Liver-Related than Alanine Aminotransferase.

OBJECTIVES: Serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) are 2 commonly ordered liver function tests, and ALT has long been considered more liver-specific than AST. Between the 2, the one which is better in predicting liver or non-liver-related mortality remains unsettled. METHODS: The cohort, 416,122 adults, came from a self-paying comprehensive health surveillance program during 1994-2008 and was followed up till 2008. Mortality came from National Death Index, with 10,412 deaths identified. Hazard ratios (HRs), computed by Cox model, and life expectancy, by life table method, were presented for 5 levels of AST and ALT with elevated AST or ALT defined as ≥40 IU/L. Liver disease included liver cancer and other liver conditions. RESULTS: There were 3 times more elevated ALT (15.4%) than AST (5.7%). However, those with elevated AST had higher mortality for all-cause (HR = 2.44), for liver disease (HR = 27.2), and for liver cancer (HR = 47.6) than its ALT counterparts (HR = 1.69, 10.8, and 20.2, respectively). Elevated AST also lost more years of life expectancy (10.2) than those lost by ALT (5.2) and larger than most common risks. Elevated AST had increased mortality from all cancers (HR = 3.57), stroke (HR = 1.36), respiratory diseases (HR = 1.34), and injuries (HR = 1.82), other than just liver disease. All-cause mortality remained significantly increased, when high risk groups were excluded, such as frequent drinkers, hepatitis carriers, those died from nonmedical conditions, those died in the first 3 years, or advanced fibrosis index based on 4 factors or aspartate transaminase-to-platelet ratio index. Results were consistent between those returned for second visits and those analyzed in initial visits. DISCUSSION: Those with elevated AST (≥40 IU/L) had life expectancy cut short by 10.2 years, doubled the number of years lost with elevated ALT. For all-cause and for liver-related mortality, AST was an important predictor, better than ALT.

Induction of squamous cell carcinoma after MAP3K8 overexpression in murine salivry gland epithelial cells.

BACKGROUND: Salivary gland neoplasms are relatively rare and comprise only 1%-4% of all human neoplasms. Salivary gland neoplasms also show an extremely wide range of morphological diversity. Currently, the genetic alterations and corresponding molecular mechanisms underlying salivary gland neoplasms development remain largely unknown. METHOD: We generated an inducible Tet-MAP3K8::MMTV-rTA mouse model by crossing the MAP3K8 transgenic mice with MMTR-rTA transgenic mice to express MAP3K8 in the salivary gland. RESULTS: MAP3K8 overexpression in the murine salivary glands of Tet-MAP3K8::MMTR-rTA transgenic mice induces tumorigenesis. Pathological investigations reveal partial fibrosis and adenosis of salivary glands, and foci of atypical squamoid cellular proliferation, which represent invasive squamous cell carcinoma (SCC). CONCLUSION: MAP3K8 overexpression is associated with SCC development in murine salivary glands. It provides an in vivo framework for the understanding of molecular mechanisms underlying SCC development in the salivary glands and also for the development of a future therapeutic strategy targeting this tumor type.

TPL2 Is an Oncogenic Driver in Keratocanthoma and Squamous Cell Carcinoma.

Squamous cell carcinoma (SCC) and keratoacanthoma (KA; SCC/KA) research has been hampered mainly by our lack of understanding the underlying genetic and epigenetic alterations associated with SCC/KA development, as well as the lack of animal models that faithfully recapitulate histopathologic features of human SCC/KA. Here, we show that TPL2 overexpression induced both cell transformation in immortalized human keratinocytes and SCC and KA-like cutaneous SCC (cSCC) development in mice. Mechanistically, activation of TPL2 downstream signaling pathways such as MEK/ERK MAPK, mTOR, NF-κB, and p38 MAPK leads to TPL2-mediated cell transformation in immortalized human keratinocytes and tumorigenesis in mice. Most importantly, TPL2 overexpression is required for iTPL2 TG-driven SCC and KA-like cSCC tumor maintenance, validating TPL2 as a possible drug target for the treatment of SCC/KA. Finally, we verified that TPL2 is overexpressed in human cutaneous metastatic SCC and KA clinical specimens compared with normal skin. Taken together, our results establish TPL2 as an oncogenic driver in SCC/KA development. Cancer Res; 76(22); 6712-22. ©2016 AACR.

Prokaryote-expressed M2e protein improves H9N2 influenza vaccine efficacy and protection against lethal influenza A virus in mice.

BACKGROUND: Influenza vaccines are prepared annually based on global epidemiological surveillance data. However, since there is no method by which to predict the influenza strain that will cause the next pandemic, the demand to develop new vaccination strategies with broad cross-reactivity against influenza viruses are clearly important. The ectodomain of the influenza M2 protein (M2e) is an attractive target for developing a vaccine with broad cross-reactivity. For these reasons, we investigated the efficacy of an inactivated H9N2 virus vaccine (a-H9N2) mixed with M2e (1xM2e or 4xM2e) proteins expressed in Escherichia coli, which contains the consensus of sequence the extracellular domain of matrix 2 (M2e) of A/chicken/Vietnam/27262/09 (H5N1) avian influenza virus, and investigated its humoral immune response and cross-protection against influenza A viruses. RESULTS: Mice were intramuscularly immunized with a-H9N2, 1xM2e alone, 4xM2e alone, a-H9N2/1xM2e, or a-H9N2/4xM2e. Three weeks post-vaccination, mice were challenged with lethal homologous (A/ chicken /Korea/ma163/04, H9N2) or heterosubtypic virus (A/Philippines/2/82, H3N2 and A/aquatic bird/Korea/maW81/05, H5N2). Our studies demonstrate that the survival of mice immunized with a-H9N2/1xM2e or with a-H9N2/4xM2e (100% survival) was significantly higher than that of mouse-adapted H9N2 virus-infected mice vaccinated with 1xM2e alone or with 4xM2e alone (0% survival). We also evaluated the protective efficacy of the M2e + vaccine against infection with mouse-adapted H5N2 influenza virus. Protection from death in the control group (0% survival) was similar to that of the 1×M2e alone and 4xM2e alone-vaccinated groups (0% survival). Only 40% of mice vaccinated with vaccine alone survived challenge with H5N2, while the a-H9N2/1×M2e and a-H9N2/4×M2e groups showed 80% and 100% survival following mouse-adapted H5N2 challenge, respectively. We also examined cross-protection against human H3N2 virus and found that the a-H9N2/1×M2e group displayed partial cross-protection against H3N2 (40% survival), whereas vaccine alone, 1×M2e alone, 4×M2e alone, or H9N2/1×M2e groups showed incomplete protection (0% survival) in response to challenge with a lethal dose of human H3N2 virus. CONCLUSIONS: Taken together, these results suggest that prokaryote-expressed M2e protein improved inactivated H9N2 virus vaccine efficacy and achieved cross-protection against lethal influenza A virus infection in mice.

Interferon regulatory factor 7 regulates glioma stem cells via interleukin-6 and Notch signalling.

Inflammatory microenvironment signalling plays a crucial role in tumour progression (i.e. cancer cell proliferation, survival, angiogenesis and metastasis) in many types of human malignancies. However, the role of inflammation in brain tumour pathology remains poorly understood. Here, we report that interferon regulatory factor 7 is a crucial regulator of brain tumour progression and heterogeneity. Ectopic expression of interferon regulatory factor 7 in glioma cells promotes tumorigenicity, angiogenesis, microglia recruitment and cancer stemness in vivo and in vitro through induction of interleukin 6, C-X-C motif chemokine 1 and C-C motif chemokine 2. In particular, interferon regulatory factor 7-driven interleukin 6 plays a pivotal role in maintaining glioma stem cell properties via Janus kinase/signal transducer and activator of transcription-mediated activation of Jagged-Notch signalling in glioma cells and glioma stem cells derived from glioma patients. Accordingly, the short hairpin RNA-mediated depletion of interferon regulatory factor 7 in glioma stem cells markedly suppressed interleukin 6-Janus kinase/signal transducer and activator of transcription-mediated Jagged-Notch-signalling pathway, leading to decreases in glioma stem cell marker expression, tumoursphere-forming ability, and tumorigenicity. Furthermore, in a mouse model of wound healing, depletion of interferon regulatory factor 7 suppressed tumour progression and decreased cellular heterogeneity. Finally, interferon regulatory factor 7 was overexpressed in patients with high-grade gliomas, suggesting its potential as an independent prognostic marker for glioma progression. Taken together, our findings indicate that interferon regulatory factor 7-mediated inflammatory signalling acts as a major driver of brain tumour progression and cellular heterogeneity via induction of glioma stem cell genesis and angiogenesis.

Role of the p21-activated kinases (PAKs) in influenza A virus replication.

Influenza A virus infection stimulates a wide range of virus-supportive or antiviral mechanisms in host cells. p21-Activated kinase 1 (PAK1) is a serine/threonine kinase that regulates a number of fundamental cellular processes and has been implicated in the modulation of virus replication. Here, we investigated the role of PAK1 activation during influenza A virus infection and found that virus propagation corresponded to stimulated PAK1 phosphorylation. Moreover, transfection of the active form of PAK1 (PAK1-T423E) in A549 cells induced higher viral titers (∼10-fold differences) compared to that in the control vector or inactive PAK1 (PAK1-K299R)-transfected cells. PAK1-specific siRNA knockdown also resulted in 10-100-fold reductions in virus yields compared to that in the control siRNA-treatment (p<0.05). We further showed that treatment with PAK18, a PAK1 peptide inhibitor, resulted in marked suppression of both ERK 1/2 phosphorylation and infectious virus production, which was comparable to that by U0126, a specific MEK/ERK inhibitor. These results provide evidence for the importance of PAK1 activation during influenza virus infection and its association with ERK in regulating virus replication. The present study also implicates PAK1 as a potential therapeutic target for managing influenza virus infections.

Impairment of the Staufen1-NS1 interaction reduces influenza viral replication.

Staufen1 (Stau1), a host cellular protein, along with non-structural protein 1 (NS1), an influenza viral protein, associate with each other during influenza viral infection and down-regulation of Stau1 by RNA interference reduces the yield of influenza A virus, suggesting a role for Stau1 in viral replication. In order to develop a new tool to control influenza A virus, we determined the specific regions of Staufen1 protein involved in the interaction with NS1. The linker between RBD3 and 4 was isolated as the binding regions. Expression of RBD3L, the linker including RBD3, inhibited the interaction between Stau1 and NS1, reducing the colocalization of the two proteins in the cytosol and nucleus regions. In addition, yield of influenza A virus in RBD3L-expressing cells was significantly reduced 36 h after infection. These results suggest that disruption of the Stau1-NS1 interaction can be used to control replication of influenza A virus, thereby providing a target for the development of antiviral drugs.

Emergence of amantadine-resistant H3N2 avian influenza A virus in South Korea.

We found a relatively high frequency of unique amantadine-resistant H3N2 and H9N2 avian influenza viruses (Val27Ile on M2 protein) isolated from live poultry markets in South Korea and confirmed that a Val27Ile single substitution in the M2 protein is enough to acquire the amantadine resistance phenotype by using reverse-genetically created human-avian reassortant viruses.

Growth inhibitory effect of triple anti-tumor gene transfer using Semliki Forest virus vector in glioblastoma cells.

The gene delivery of multiple tumor suppressors can provide an efficient tumor therapy in the case of malignant human glioblastomas containing multiple genetic alteration and inactivation. As such, the current study presents a new delivery system that can simultaneously express three anti-tumor genes using a Semliki Forest virus (SFV) vector in the expectation of combined or synergistic effects of angiogenesis inhibition by angiostatin and apoptosis induction by p53, PTEN and the rSFV particle itself. Recombinant SFV (rSFV) containing three anti-tumor genes (rSFV-Agt/p53/PTEN) were found to efficiently transduce and express each anti-tumor gene in glioblastoma cells. In addition, rSFV-Agt/p53/PTEN also resulted in a more effective induction of apoptosis in vitro and inhibition of tumor growth in nude mice when compared with other rSFVs containing only one or two anti-tumor genes. Accordingly, the current results demonstrate that a triple anti-tumor gene transfer using an rSFV vector would be a powerful strategy for regional cancer gene therapy.