Search

Purpose: In a recent phase II study of onartuzumab (MetMAb), patients whose non–small cell lung cancer (NSCLC) tissue scored as positive for MET protein by immunohistochemistry (IHC) experienced a significant benefit with onartuzumab plus erlotinib (O+E) versus erlotinib. We describe development and validation of a standardized MET IHC assay and, retrospectively, evaluate multiple biomarkers as predictors of patient benefit.Experimental Design: Biomarkers related to MET and/or EGF receptor (EGFR) signaling were measured by IHC, FISH, quantitative reverse transcription PCR, mutation detection techniques, and ELISA.Results: A positive correlation between IHC, Western blotting, and MET mRNA expression was observed in NSCLC cell lines/tissues. An IHC scoring system of MET expression taking proportional and intensity-based thresholds into consideration was applied in an analysis of the phase II study and resulted in the best differentiation of outcomes. Further analyses revealed a nonsignificant overall survival (OS) improvement with O+E in patients with high MET copy number (mean ≥5 copies/cell by FISH); however, benefit was maintained in “MET IHC-positive”/MET FISH-negative patients (HR, 0.37; P = 0.01). MET, EGFR, amphiregulin, epiregulin, or HGF mRNA expression did not predict a significant benefit with onartuzumab; a nonsignificant OS improvement was observed in patients with high tumor MET mRNA levels (HR, 0.59; P = 0.23). Patients with low baseline plasma hepatocyte growth factor (HGF) exhibited an HR for OS of 0.519 (P = 0.09) in favor of onartuzumab treatment.Conclusions: MET IHC remains the most robust predictor of OS and progression-free survival benefit from O+E relative to all examined exploratory markers. Clin Cancer Res; 20(17); 4488–98. ©2014 AACR.

Additional methodology describing cell lines and tissues, as well as detailed methodology for gene expression, flow cytometry, IHC and FISH studies.

Figure S1. Characterization of SP44 specificity; Figure S2. Relationship of IHC intensity with FACS; Figure S3. Relationship of IHC intensity with mRNA; Figure S4. Representative IHC images; Figure S5 and S6. MET IHC H-scores; Figure S7. Relationship of IHC clinical score with MET copy number; Figure S8. Kaplan-Meier curves of OS according to MET copy number; Figure S9. Relationship of plasma HGF with MET Dx positivity.

Table S1. Overlap of biomarkers measured in OAM4558g; Table S2. Primer/probe information; Table S3. MET diagnostic scoring criteria; Table S4. MET IHC intratumoral heterogeneity; Table S5. Impact of KRAS/EGFR mutation on clinical outcome; Table S6. Biomarker status in EGFR-mutant patients; Table S7. Association of transcript expression with PFS in OAM4558g.

Breast Ultrasound (BUS) imaging has been recognized as an essential imaging modality for breast masses classification in China. Current deep learning (DL) based solutions for BUS classification seek to feed ultrasound (US) images into deep convolutional neural networks (CNNs), to learn a hierarchical combination of features for discriminating malignant and benign masses. One existing problem in current DL-based BUS classification was the lack of spatial and channel-wise features weighting, which inevitably allow interference from redundant features and low sensitivity. In this study, we aim to incorporate the instructive information provided by breast imaging reporting and data system (BI-RADS) within DL-based classification. A novel DL-based BI-RADS Vector-Attention Network (BVA Net) that trains with both texture information and decoded information from BI-RADS stratifications was proposed for the task. Three baseline models, pre-trained DenseNet-121, ResNet-50 and Residual-Attention Network (RA Net) were included for comparison. Experiments were conducted on a large scale private main dataset and two public datasets, UDIAT and BUSI. On the main dataset, BVA Net outperformed other models, in terms of AUC (area under the receiver operating curve, 0.908), ACC (accuracy, 0.865), sensitivity (0.812) and precision (0.795). BVA Net also achieved the high AUC (0.87 and 0.882) and ACC (0.859 and 0.843), on UDIAT and BUSI. Moreover, we proposed a method that integrates both BVA Net binary classification and BI-RADS stratification estimation, called integrated classification. The introduction of integrated classification helped improving the overall sensitivity while maintaining a high specificity.

Travel, as one way to relax oneself, has become the first choice for people to enjoy their body and mind in modern society. However, while facing lots of information, how to help users make better decisions on their next travel goals through their historical interest spots is a direction that needs further research in big data recommendation systems. In this thesis, we proposed the deep convolution and multi-head self-attention position network model. First, it extracts the user’s historical interest point feature information by convolutional neural network method, and then performs horizontal and vertical filtering. Next, it interacts the obtained information with the candidate attraction information, and extracts the location information of the historical interest sequence by the multi-head self-attention mechanism. Finally, the model does the attention mechanism of the candidate attraction by fusing the feature information of the location information. The final model achieves a deep fusion of user sequence interest and location feature information. We conducted detailed comparison experiments with the very popular models in the industry on different public datasets, and the results showed that our deep convolution and multi-head self-attention position network model has good performance.

Tissue regeneration and wound healing are still serious clinical complications globally and lack satisfactory cures. Inspired by the impressive regeneration ability of the post-injury earthworms and their widely accepted medicinal properties, we screened and identified a novel collagen-like peptide from the amputated earthworms using high-throughput techniques, including transcriptomics, proteomics, and mass spectrum. The identified collagen-like peptide col4a1 was cloned and expressed to comprehensively investigate the wound healing effect and underlying mechanism. It exerted significant effects on wound healing both in vitro and in vivo, including enhanced viability, proliferation, migration of fibroblasts, granulation, and collagen deposition. Moreover, the col4a1 functioned via binding with integrin α2β1 and upregulating the RAS/MAPK signaling pathway. This work demonstrates that the novel collagen-like peptide col4a1 obtained from the amputated earthworms enables enhanced wound healing and provides new opportunities for wound care.

Background: To investigate deep vein thrombosis (DVT) in hospitalized patients with coronavirus disease 2019 (COVID-19), we performed a single institutional study to evaluate its prevalence, risk factors, prognosis, and potential thromboprophylaxis strategies in a large referral and treatment center. Methods: We studied a total of 143 patients with COVID-19 from January 29, 2020 to February 29, 2020. Demographic and clinical data, laboratory data, including ultrasound scans of the lower extremities, and outcome variables were obtained, and comparisons were made between groups with and without DVT. Results: Of the 143 patients hospitalized with COVID-19 (age 63±14 years, 74 [51.7%] men), 66 patients developed lower extremity DVT (46.1%: 23 [34.8%] with proximal DVT and 43 [65.2%] with distal DVT). Compared with patients who did not have DVT, patients with DVT were older and had a lower oxygenation index, a higher rate of cardiac injury, and worse prognosis, including an increased proportion of deaths (23 [34.8%] versus 9 [11.7%]; P =0.001) and a decreased proportion of patients discharged (32 [48.5%] versus 60 [77.9%]; P P =0.031), Padua prediction score ≥4 (odds ratio, 4.016; P =0.04), D-dimer >1.0 μg/mL (odds ratio, 5.818; P 1.0 μg/mL has a sensitivity of 88.52% and a specificity of 61.43% for screening for DVT. In the subgroup of patients with a Padua prediction score ≥4 and whose ultrasound scans were performed >72 hours after admission, DVT was present in 18 (34.0%) patients in the subgroup receiving venous thromboembolism prophylaxis versus 35 (66.0%) patients in the nonprophylaxis group ( P =0.010). Conclusions: The prevalence of DVT is high and is associated with adverse outcomes in hospitalized patients with COVID-19. Prophylaxis for venous thromboembolism may be protective in patients with a Padua protection score ≥4 after admission. Our data seem to suggest that COVID-19 is probably an additional risk factor for DVT in hospitalized patients.

JNK signaling plays a critical role in both tumor promotion and tumor suppression. Here, we identified clustered microRNAs (miRNAs) miR-306 and miR-79 as novel tumor-suppressor miRNAs that specifically eliminate JNK-activated tumors in Drosophila. While showing no significant effect on normal tissue growth, miR-306 and miR-79 strongly suppressed growth of multiple tumor models including malignant tumors caused by Ras activation and cell polarity defects. Mechanistically, these miRNAs commonly target the mRNA of an E3 ubiquitin ligase Drosophila ring finger protein 146 (dRNF146). We found that DRNF146 promotes degradation of tankyrase (Tnks), an ADP-ribose polymerase that promotes JNK activation in a non-canonical manner. Thus, downregulation of dRNF146 by miR-306 and miR-79 leads to hyper-enhancement of JNK activation. Our data show that, while JNK activity is essential for tumor growth, elevation of miR-306 or miR-79 overactivate JNK signaling to the lethal level via non-canonical JNK pathway and thus eliminate tumors, providing a new miRNA-based strategy against cancer.

In recent years, the problem of youth unemployment in China has become a great concern. Unemployed young people often find themselves under tremendous stress and vulnerable to sleep problems. The link between perceived stress and insomnia has been widely documented, but the potential mechanisms underlying this relationship have not been thoroughly revealed. The objective of this study was to examine the underlying mechanisms linking perceived stress to insomnia through testing the mediating effects of emotional dysregulation and time structure. Through a multiple-stage convenience sampling, 511 unemployed young people (38.6% women; mean age = 21.51; SD = 2.22) were measured using the 10-item Perceived Stress Scale (PSS-10), Insomnia Severity Index (ISI), 16-item Difficulties in Emotion Regulation Scale (DERS-16), and Time Structure Questionnaire (TSQ). Based on the statistical description of the sample, chi-squared tests, bivariate correlation analyses, and mediation analyses were performed. The study indicated that 53.0% (95% CI: 48.7–57.4%) of the participants reported probable insomnia, and both insomnia and perceived stress demonstrated strong associations with emotional dysregulation and time structure, which served as partial mediators between perceived stress and insomnia symptoms according to mediation analyses. Path analysis further revealed that, after controlling for age and gender, emotional dysregulation and time structure accounted for 31.8 and 17.6% of the effect of perceived stress on insomnia, respectively. This study provides empirical support for the association among perceived stress, emotional dysregulation, time structure, and insomnia symptoms. To improve the sleep quality and general wellbeing of unemployed young people, emotional dysregulation and time structure must be taken into consideration.

Whether in the field of personalized advertising or recommender systems, click through rate (CTR) prediction is a very important task. In recent years, Alibaba Group has done a lot of advanced research on the prediction of click through rate, and proposed a technical route includes kinds of deep learning models. For example, in the deep interest network (DIN) proposed by the Alibaba Group, the sequence of users’ browsing behaviors is used to express their interest features, and this sequence is made up of items clicked by users. Usually, the item is mapped into a static vector, but the fixed length embedded vector is difficult to express the user’s dynamic interest features. In order to solve this problem, Alibaba Group introduced the attention mechanism in the field of natural language processing (NLP) into deep interest network (DIN), and designed a unique activation unit to extract the important informations in the user’s historical behavior sequence, and use these important informations to express user’s dynamic interest features. In this paper, we propose a novel deep learning model: Multi-Interest Network Based on Double Attention for Click-Through Rate Prediction (DAMIN), which based on Deep Interest Network (DIN) and combined with multi-head attention mechanism. In the deep interest network (DIN), the attention weight between the candidate item vector and the item vector of the user’s historical behavior sequence is learned by fully connected neural network. Different from deep interest network (DIN), we design a new method, which uses the reciprocal of Euclidean Distance to represent the attention weight between two item vectors. Then, the item vectors in user’s historical behavior sequence are weighted by the attention weights and meanwhile the candidate item vectors are also weighted by the attention weights. In the next, we can obtain new item vectors by add the weighted item vectors of user’s historical behavior sequence and weighted candidate item vectors, and those new item vectors are used to represent user’s dynamic interest feature vectors. In the end, the user’s dynamic interest features are send into the three multi-head attention layers, which can extract users’ various interest features. We have conducted a lot of experiments on three real-world datasets of Amazon and the results show that the model proposed by this paper acquires a better performance than some classical models. Compared with DIN, the model proposed in this paper improves the average of AUC by 4%–5%, which proves that the model proposed in this paper is effective. In addition, a large number of ablation experiments have been carried out to prove that each module of the proposed model is effective.

Immunohistochemistry (IHC) assays play a central role in evaluating biomarker expression in tissue sections for diagnostic and research applications. Manual scoring of IHC images, which is the current standard of practice, is known to have several shortcomings in terms of reproducibility and scalability to large scale studies. Here, by using a digital image analysis-based approach, we introduce a new metric called the pixelwise H-score (pix H-score) that quantifies biomarker expression from whole-slide scanned IHC images. The pix H-score is an unsupervised algorithm that only requires the specification of intensity thresholds for the biomarker and the nuclear-counterstain channels. We present the detailed implementation of the pix H-score in two different whole-slide image analysis software packages Visiopharm and HALO. We consider three biomarkers P-cadherin, PD-L1, and 5T4, and show how the pix H-score exhibits tight concordance to multiple orthogonal measurements of biomarker abundance such as the biomarker mRNA transcript and the pathologist H-score. We also compare the pix H-score to existing automated image analysis algorithms and demonstrate that the pix H-score provides either comparable or significantly better performance over these methodologies. We also present results of an empirical resampling approach to assess the performance of the pix H-score in estimating biomarker abundance from select regions within the tumor tissue relative to the whole tumor resection. We anticipate that the new metric will be broadly applicable to quantify biomarker expression from a wide variety of IHC images. Moreover, these results underscore the benefit of digital image analysis-based approaches which offer an objective, reproducible, and highly scalable strategy to quantitatively analyze IHC images.

Recently, deep learning has been widely used in the field of recommendation systems. Sequence recommendation is an important direction of the current recommendation system, which uses the user's behavior sequence information to calculate the possibility of the user interacting with the target item. Recent work embeds the original features into low-dimensional vectors, then uses RNN or Transformer to extract behavior sequence information, and finally inputs to the MLP layer to get the recommended results. These methods do not consider that the items of different positions have different effects on the next item, and the recent item in the real world often have a greater impact on the target item than the previous ones. So we propose MAUPRec models to solve such problems. Each user's historical behavior learns its corresponding weight through attention mechanism, which better reflects the interest of users at different times. In addition, we find that feedback information means user's preference to some extent, so we also introduce feedback information as user's preference representation in the model. We conducted detailed comparison experiments with the very popular models in the industry on different public data sets, and the results showed that our model MAUPRec has good performance.

Evidence is limited regarding whether ambient monitoring can properly represent personal ozone exposure. We conducted a longitudinal panel study to measure personal exposure to ozone using real-time personal ozone monitors. Corresponding ambient ozone concentrations and possible influencing factors (meteorological conditions and activity patterns) were also collected. We used linear mixed-effect models to analyze personal-ambient ozone concentration associations and possible influencing factors. Ambient ozone concentrations were around two to three times higher than personal ozone (43.1 μg/m3 on average) and their correlations were weak with small slopes (0.35) and marginal R square (RM2) values (0.24). Larger RM2 values were found under high temperature (>29.5 °C), low humidity (4 h) and for individuals spent longer time outdoors (>0.6 h). In final model, personal ozone exposure was positively associated with ambient concentrations and ventilation conditions, but inversely correlated with ambient temperature and humidity. The models explained >50% of personal ozone concentration variabilities. Our results highlight that ambient ozone concentration alone is not a suitable surrogate for individual exposure assessment. Meteorological conditions (temperature and humidity) and activity patterns (windows opening and outdoor activities) that affecting personal ozone exposure should be taken into account. Keywords: Ozone, Personal exposure, Ambient concentrations, Influencing factors, Panel study

Graph-based malware detection methods must build a behavior graph for each known malware, and they are difficult to apply in practice. To solve this issue, we study how to build a common behavior graph for each malware family. We represent malware behaviors as dependency graphs. To find the dependency relations between system calls, we use a dynamic taint analysis technique to mark the system call parameters with taint tags, and we then build the system call dependency graph by tracing the propagation of the taint data. Based on the dependency graphs of malware samples, we propose an algorithm to extract the common behavior graph, which is used to represent the behavioral features of a malware family. Finally, a graph matching algorithm that is based on the maximum weight subgraph is used to detect malicious code. The experimental results show that the proposed method has a high detection rate and a low false positive rate and can detect malware variants.