Search

Table S1:Summary of baseline pathological diagnosis; Table S2:Noncompartmental Pharmacokinetic Summary Following Oral Administration of ralimetinib on Day 1 and on Day 14 (Cycle 1) - Capsule Formulation; Table S3: Noncompartmental Pharmacokinetic Summary Following Oral Administration of ralimetinib on Day 1 and on Day 14 (Cycle 1) - Tablet Formulation

Trial Profile

Purpose: p38 MAPK regulates the production of cytokines in the tumor microenvironment and enables cancer cells to survive despite oncogenic stress, radiotherapy, chemotherapy, and targeted therapies. Ralimetinib (LY2228820 dimesylate) is a selective small-molecule inhibitor of p38 MAPK. This phase I study aimed to evaluate the safety and tolerability of ralimetinib, as a single agent and in combination with tamoxifen, when administered orally to patients with advanced cancer.Experimental Design: The study design consisted of a dose-escalation phase performed in a 3+3 design (Part A; n = 54), two dose-confirmation phases [Part B at 420 mg (n = 18) and Part C at 300 mg (n = 8)], and a tumor-specific expansion phase in combination with tamoxifen for women with hormone receptor–positive metastatic breast cancer refractory to aromatase inhibitors (Part D; n = 9). Ralimetinib was administered orally every 12 hours on days 1 to 14 of a 28-day cycle.Results: Eighty-nine patients received ralimetinib at 11 dose levels (10, 20, 40, 65, 90, 120, 160, 200, 300, 420, and 560 mg). Plasma exposure of ralimetinib (Cmax and AUC) increased in a dose-dependent manner. After a single dose, ralimetinib inhibited p38 MAPK–induced phosphorylation of MAPKAP-K2 in peripheral blood mononuclear cells. The most common adverse events, possibly drug-related, included rash, fatigue, nausea, constipation, pruritus, and vomiting. The recommended phase II dose was 300 mg every 12 hours as monotherapy or in combination with tamoxifen. Although no patients achieved a complete response or partial response,19 patients (21.3%) achieved stable disease with a median duration of 3.7 months, with 9 of these patients on study for ≥6 cycles.Conclusions: Ralimetinib demonstrated acceptable safety, tolerability, and pharmacokinetics for patients with advanced cancer. Clin Cancer Res; 22(5); 1095–102. ©2015 AACR.

MotivationSingle-cell RNA-sequencing (scRNA-seq) technology enables researchers to investigate a genome at the cellular level with unprecedented resolution. An organism consists of a heterogeneous collection of cell types, each of which plays a distinct role in various biological processes. Hence, the first step of scRNA-seq data analysis is often to distinguish cell types so they can be investigated separately. Researchers have recently developed several automated cell-type annotation tools, requiring neither biological knowledge nor subjective human decisions. Dropout is a crucial characteristic of scRNA-seq data widely used in differential expression analysis. However, no current cell annotation method explicitly utilizes dropout information. Fully utilizing dropout information motivated this work.ResultsWe present scAnnotate, a cell annotation tool that fully utilizes dropout information. We model every gene’s marginal distribution using a mixture model, which describes both the dropout proportion and the distribution of the non-dropout expression levels. Then, using an ensemble machine learning approach, we combine the mixture models of all genes into a single model for cell-type annotation. This combining approach can avoid estimating numerous parameters in the high-dimensional joint distribution of all genes. Using 14 real scRNA-seq datasets, we demonstrate that scAnnotate is competitive against nine existing annotation methods. Furthermore, because of its distinct modelling strategy, scAnnotate’s misclassified cells differ greatly from competitor methods. This suggests using scAnnotate together with other methods could further improve annotation accuracy.Availability and implementationWe implemented scAnnotate as an R package and made it publicly available from CRAN: https://cran.r-project.org/package=scAnnotate.Supplementary informationSupplementary data are available at Bioinformatics Advances online.

Deep Learning (DL) has been broadly applied to solve big data problems in biomedical fields, which is most successful in image processing. Recently, many DL methods have been applied to analyze genomic studies. However, genomic data usually has too small a sample size to fit a complex network. They do not have common structural patterns like images to utilize pre-trained networks or take advantage of convolution layers. The concern of overusing DL methods motivates us to evaluate DL methods’ performance versus popular non-deep Machine Learning (ML) methods for analyzing genomic data with a wide range of sample sizes. In this paper, we conduct a benchmark study using the UK Biobank data and its many random subsets with different sample sizes. The original UK Biobank data has about 500k participants. Each patient has comprehensive patient characteristics, disease histories, and genomic information, i.e., the genotypes of millions of Single-Nucleotide Polymorphism (SNPs). We are interested in predicting the risk of three lung diseases: asthma, COPD, and lung cancer. There are 205,238 participants have recorded disease outcomes for these three diseases. Five prediction models are investigated in this benchmark study, including three non-deep machine learning methods (Elastic Net, XGBoost, and SVM) and two deep learning methods (DNN and LSTM). Besides the most popular performance metrics, such as the F1-score, we promote the hit curve, a visual tool to describe the performance of predicting rare events. We discovered that DL methods frequently fail to outperform non-deep ML in analyzing genomic data, even in large datasets with over 200k samples. The experiment results suggest not overusing DL methods in genomic studies, even with biobank-level sample sizes. The performance differences between DL and non-deep ML decrease as the sample size of data increases. This suggests when the sample size of data is significant, further increasing sample sizes leads to more performance gain in DL methods. Hence, DL methods could be better if we analyze genomic data bigger than this study.

Classification tree is a widely used machine learning method. It has multiple implementations as R packages, rpart, ctree, evtree, tree and C5.0. The details of these implementations are not the same, and hence their performances differ from one application to another. We are interested in their performance in the classification of cells using the single-cell RNA-Sequencing data. In this paper, we conducted a benchmark study using 22 Single-Cell RNA-sequencing data sets. Using cross-validation, we compare packages’ prediction performances based on their Precision, Recall, F1-score, Area Under the Curve (AUC). We also compared the Complexity and Run-time of these R packages. Our study shows that rpart and evtree have the best Precision, evtree is the best in Recall, F1-score and AUC, C5.0 prefers more complex trees, tree is consistently much faster than others, although its complexity is often higher than others.

Survival analysis is a technique for identifying prognostic biomarkers and genetic vulnerabilities in cancer studies. Large-scale consortium-based projects have profiled >11 000 adult and >4000 pediatric tumor cases with clinical outcomes and multiomics approaches. This provides a resource for investigating molecular-level cancer etiologies using clinical correlations. Although cancers often arise from multiple genetic vulnerabilities and have deregulated gene sets (GSs), existing survival analysis protocols can report only on individual genes. Additionally, there is no systematic method to connect clinical outcomes with experimental (cell line) data. To address these gaps, we developed cSurvival (https://tau.cmmt.ubc.ca/cSurvival). cSurvival provides a user-adjustable analytical pipeline with a curated, integrated database and offers three main advances: (i) joint analysis with two genomic predictors to identify interacting biomarkers, including new algorithms to identify optimal cutoffs for two continuous predictors; (ii) survival analysis not only at the gene, but also the GS level; and (iii) integration of clinical and experimental cell line studies to generate synergistic biological insights. To demonstrate these advances, we report three case studies. We confirmed findings of autophagy-dependent survival in colorectal cancers and of synergistic negative effects between high expression of SLC7A11 and SLC2A1 on outcomes in several cancers. We further used cSurvival to identify high expression of the Nrf2-antioxidant response element pathway as a main indicator for lung cancer prognosis and for cellular resistance to oxidative stress-inducing drugs. Altogether, these analyses demonstrate cSurvival’s ability to support biomarker prognosis and interaction analysis via gene- and GS-level approaches and to integrate clinical and experimental biomedical studies.

Genome-wide association study (GWAS) is a popular approach to investigate relationships between genetic information and diseases. A number of associations are tested in a study and the results are often corrected using multiple adjustment methods. It is observed that GWAS studies suffer adequate statistical power for reliability. Hence, we document known models for reliability assessment using improved statistical power in GWAS analysis.

We study the impact of capital market openness on high‐frequency market quality in China. The Shanghai–Hong Kong Stock Connect program (SHHKConnect) opens China's stock market to foreign investors and offers a natural experiment to investigate this question. Using a difference‐in‐differences approach, we find that market liberalization leads to lower quoted spread, lower effective spread, lower market depth, and higher short‐term volatility. Our findings imply that opening the markets to more sophisticated foreign investors is associated with higher competition and more cross‐market arbitrage activities, narrowing the spread and reducing liquidity providers’ profits, but increasing the price impact and short‐term volatility of connected stocks.

The new technology of single-cell RNA sequencing (scRNA-seq) can yield valuable insights into gene expression and give critical information about the cellular compositions of complex tissues. In recent years, vast numbers of scRNA-seq datasets have been generated and made publicly available, and this has enabled researchers to train supervised machine learning models for predicting or classifying various cell-level phenotypes. This has led to the development of many new methods for analyzing scRNA-seq data. Despite the popularity of such applications, there has as yet been no systematic investigation of the performance of these supervised algorithms using predictors from various sizes of scRNA-seq datasets. In this study, 13 popular supervised machine learning algorithms for cell phenotype classification were evaluated using published real and simulated datasets with diverse cell sizes. This benchmark comprises two parts. In the first, real datasets were used to assess the computing speed and cell phenotype classification performance of popular supervised algorithms. The classification performances were evaluated using the area under the receiver operating characteristic curve, F1-score, Precision, Recall, and false-positive rate. In the second part, we evaluated gene-selection performance using published simulated datasets with a known list of real genes. The results showed that ElasticNet with interactions performed the best for small and medium-sized datasets. The NaiveBayes classifier was found to be another appropriate method for medium-sized datasets. With large datasets, the performance of the XGBoost algorithm was found to be excellent. Ensemble algorithms were not found to be significantly superior to individual machine learning methods. Including interactions in the ElasticNet algorithm caused a significant performance improvement for small datasets. The linear discriminant analysis algorithm was found to be the best choice when speed is critical; it is the fastest method, it can scale to handle large sample sizes, and its performance is not much worse than the top performers.

MotivationSingle-cell RNA-sequencing (scRNA-seq) technology enables researchers to investigate a genome at the cellular level with unprecedented resolution. An organism consists of a heterogeneous collection of cell types, each of which plays a distinct role in various biological processes. Hence, the first step of scRNA-seq data analysis is often to distinguish cell types so they can be investigated separately. Researchers have recently developed several automated cell type annotation tools, requiring neither biological knowledge nor subjective human decisions. Dropout is a crucial characteristic of scRNA-seq data widely used in differential expression analysis. However, dropout information is not explicitly used by any current cell annotation method. Fully utilizing dropout information for cell type annotation motivated this work.ResultsWe present scAnnotate, a cell annotation tool that fully utilizes dropout information. We model every gene’s marginal distribution using a mixture model, which describes both the dropout proportion and the distribution of the non-dropout expression levels. Then, using an ensemble machine learning approach, we combine the mixture models of all genes into a single model for cell-type annotation. This combining approach can avoid estimating numerous parameters in the high-dimensional joint distribution of all genes. Using fourteen real scRNA-seq datasets, we demonstrate that scAnnotate is competitive against nine existing annotation methods. Furthermore, because of its distinct modelling strategy, scAnnotate’s misclassified cells are very different from competitor methods. This suggests using scAnnotate together with other methods could further improve annotation accuracy.AvailabilityWe implemented scAnnotate as an R package and made it publicly available from CRAN.ContactXuekui Zhang: xuekui@uvic.ca and Li Xing: li.xing@math.usask.ca

Survival analysis is a technique to identify prognostic biomarkers and genetic vulnerabilities in cancer studies. Large-scale consortium-based projects have profiled >11,000 adult and >4,000 paediatric tumor cases with clinical outcomes and multi-omics approaches. This provides a resource for investigating molecular-level cancer etiologies using clinical correlations. Although cancers often arise from multiple genetic vulnerabilities and have deregulated gene sets (GSs), existing survival analysis protocols can report only on individual genes. Additionally, there is no systematic method to connect clinical outcomes with experimental (cell line) data. To address these gaps, we developed cSurvival (https://tau.cmmt.ubc.ca/cSurvival). cSurvival provides a user-adjustable analytical pipeline with a curated, integrated database, and offers three main advances: (a) joint analysis with two genomic predictors to identify interacting biomarkers, including new algorithms to identify optimal cutoffs for two continuous predictors; (b) survival analysis not only at the gene, but also the GS level; and (c) integration of clinical and experimental cell line studies to generate synergistic biological insights. To demonstrate these advances, we report three case studies. We confirmed findings of autophagy-dependent survival in colorectal cancers and of synergistic negative effects between high expression of SLC7A11 and SLC2A1 on outcomes in several cancers. We further used cSurvival to identify high expression of the Nrf2-antioxidant response element pathway as a main indicator for lung cancer prognosis and for cellular resistance to oxidative stress-inducing drugs. Together, these analyses demonstrate cSurvival’s ability to support biomarker prognosis and interaction analysis via gene- and GS-level approaches and to integrate clinical and experimental biomedical studies.Key pointsWe developed cSurvival, an advanced framework using clinical correlations to study biomarker interactions in cancers, with source code and curated datasets freely available for allcSurvival includes new algorithms to identify optimal cutoffs for two continuous predictors to stratify patients into risk groups, enabling for the first time joint analysis with two genomic predictors;cSurvival allows survival analysis at the gene set (GS) level with comprehensive and up-to-date GS librariesThe cSurvival pipeline integrates clinical outcomes and experimental cancer cell line data to generate synergistic biological insights and to mine for appropriate preclinical cell line toolscSurvival is built on a manually curated cancer outcomes database

BackgroundCOVID-19 is a highly transmissible infectious disease that has infected over 122 million individuals worldwide. To combat this pandemic, governments around the world have imposed lockdowns. However, the impact of these lockdowns on the rates of COVID-19 transmission in communities is not well known. Here, we used COVID-19 case counts from 3,000+ counties in the United States (US) to determine the relationship between lockdown as well as other county factors and the rate of COVID-19 spread in these communities.MethodsWe merged county-specific COVID-19 case counts with US census data and the date of lockdown for each of the counties. We then applied a Functional Principal Component (FPC) analysis on this dataset to generate scores that described the trajectory of COVID-19 spread across the counties. We used machine learning methods to identify important factors in the county including the date of lockdown that significantly influenced the FPC scores.FindingsWe found that the first FPC score accounted for up to 92.81% of the variations in the absolute rates of COVID-19 as well as the topology of COVID-19 spread over time at a county level. The relation between incidence of COVID-19 and time at a county level demonstrated a hockey-stick appearance with an inflection point approximately 7 days prior to the county reporting at least 5 new cases of COVID-19; beyond this inflection point, there was an exponential increase in incidence. Among the risk factors, lockdown and total population were the two most significant features of the county that influenced the rate of COVID-19 infection, while the median family income, median age and within-county move also substantially affect COVID spread.InterpretationLockdowns are an effective way of controlling the COVID-19 spread in communities. However, significant delays in lockdown cause a dramatic increase in the case counts. Thus, the timing of the lockdown relative to the case count is an important consideration in controlling the pandemic in communities.Research in contextEvidence before this studyWe searched PubMed using the term “coronavirus”, OR “COVID-19”, OR “COVID-19 infection”, OR “SARS-CoV-2” combined with “Lockdown” or “sociodemographic factor” or “Vulnerability” for original articles published before March 18, 2021. Similar searches were done in medRxiv, Google Scholar, and Web of Science. Only papers published in English were reviewed. The most similar relevant works to our study were Acharya et al.1 and Karmakar et al.2, which investigated the associations between population-level social factors and COVID-19 incidence and mortality. Unlike our current study, which employed a longitudinal design, both of studies were cross-sectional in nature and thus fixed on a single time point. In addition, neither of these studies investigated the impact of lockdown measures on COVID-19 infection patterns. Another relevant study is Alfano et al.’s work3, which focused on the efficacy of lockdown on COVID-19 case rates. However, this study did not evaluate the timing of lockdown on this endpoint.Added value of this studyTo our knowledge, this is the first study to use functional principal component analysis (FPCA) to investigate COVID-19 infection trajectories (in a longitudinal manner) and their relationships with different sociodemographic factors and lockdown policy at a county level. The FPCA transformed a longitudinal vector with high-dimensions into a “single” surrogate variable, which retained 93% of the information. We used an advanced statistical model (segmented regression) to investigate the effects of lockdown on incidence of COVID-19 across the US. We found that the relationship had a “hockey stick” appearance with an inflection point at ∼7 days prior to a county reporting at least 5 cases of COVID-19. We also applied a machine learning model (i.e., elastic net) to explore joint effects of lockdown and other sociodemographic factors on COVID-19 infection patterns, which estimated the impact of each of factors, adjusted for each other.Implications of all the available evidenceOur study suggests that lockdown is an effective policy to reduce case counts of COVID-19 in communities; however, significant delays in its implementation result in exponential growth of COVID-19. The inflection point is approximately 7 days prior to a county reporting at least 5 cases of COVID-19. These data will help policy-makers to determine the optimal timing of lockdowns for their communities.

Background: Single-cell RNA sequencing (scRNA-seq) yields valuable insights about gene expression and gives critical information about complex tissue cellular composition. In the analysis of single-cell RNA sequencing, the annotations of cell subtypes are often done manually, which is time-consuming and irreproducible. Garnett is a cell-type annotation software based the on elastic net method. Besides cell-type annotation, supervised machine learning methods can also be applied to predict other cell phenotypes from genomic data. Despite the popularity of such applications, there is no existing study to systematically investigate the performance of those supervised algorithms in various sizes of scRNA-seq data sets. Methods and Results: This study evaluates 13 popular supervised machine learning algorithms to classify cell phenotypes, using published real and simulated data sets with diverse cell sizes. The benchmark contained two parts. In the first part, we used real data sets to assess the popular supervised algorithms' computing speed and cell phenotype classification performance. The classification performances were evaluated using AUC statistics, F1-score, precision, recall, and false-positive rate. In the second part, we evaluated gene selection performance using published simulated data sets with a known list of real genes. Conclusion: The study outcomes showed that ElasticNet with interactions performed best in small and medium data sets. NB was another appropriate method for medium data sets. In large data sets, XGB works excellent. Ensemble algorithms were not significantly superior to individual machine learning methods. Adding interactions to ElasticNet can help, and the improvement was significant in small data sets., Comment: 21 pages, 4 figures, 1 table

Background: COVID-19 is a highly transmissible infectious disease that has infected over 122 million individuals worldwide. To combat this pandemic, governments around the world have imposed lockdowns. However, the impact of these lockdowns on the rates of COVID-19 transmission in communities is not well-known. Here, we used COVID-19 case counts from 3,000+ counties in the United States (US) to determine the relationship between lockdown as well as other county factors and the rate of COVID-19 spread in these communities. Methods: We merged county-specific COVID-19 case counts with US census data and the date of lockdown for each of the counties. We then applied a Functional Principal Component (FPC) analysis on this dataset to generate scores that described the trajectory of COVID-19 spread across the counties. We used machine learning methods to identify important factors in the county including the date of lockdown that significantly influenced the FPC scores. Findings: We found that the first FPC score accounted for up to 92.81% of the variations in the absolute rates and the topology of COVID-19 spread over time at a county level. The relation between incidence of COVID-19 and time at a county level demonstrated a hockey-stick appearance with an inflection point approximately 7 days prior to the county reporting at least 5 new cases of COVID-19; beyond this inflection point, there was an exponential increase in incidence. Among the risk factors, lockdown and total population were the two most significant features of the county that influenced the rate of COVID-19 infection, while the median family income, median age and within-county move also substantially affect COVID spread. Interpretation: Lockdowns are an effective way of controlling the COVID-19 spread in communities. However, significant delays in lockdown cause a dramatic increase in the case counts. Thus, the timing of the lockdown relative to the case count is an important consideration in controlling the pandemic in communities. Funding Statement: Dr. Xuekui Zhang is funded by Canada Research Chairs. Grant Number: 950-231363 and Natural Sciences and Engineering Research Council of Canada. Grant Number: RGPIN-2017-04722. This research was enabled in part by support provided by WestGrid (www.westgrid.ca) and Compute Canada (www.computecanada.ca). The computing resource is provided by Compute Canada Resource Allocation Competitions #3495 (PI: Xuekui Zhang) and #1551 (PI: Li Xing). Dr. Don Sin is a Tier 1 Canada Research Chair in COPD and holds the de Lazzari Family Chair at the Heart Lung Innovation, Vancouver, Canada. Declaration of Interests: Don Sin: Professor Sin reports grants from Merck, personal fees from Sanofi-Aventis, personal fees from Regeneron, grants and personal fees from Boehringer Ingelheim, grants and personal fees from AstraZeneca, personal fees from Novartis, outside the submitted work. Other coauthors have nothing to declare.

As a future trend of healthcare, personalized medicine tailors medical treatments to individual patients. It requires to identify a subset of patients with the best response to treatment. The subset can be defined by a biomarker (e.g. expression of a gene) and its cutoff value. Topics on subset identification have received massive attention. There are over 2 million hits by keyword searches on Google Scholar. However, how to properly incorporate the identified subsets/biomarkers to design clinical trials is not trivial and rarely discussed in the literature, which leads to a gap between research results and real-world drug development. To fill in this gap, we formulate the problem of clinical trial design into an optimization problem involving high-dimensional integration, and propose a novel computational solution based on Monte-Carlo and smoothing methods. Our method utilizes the modern techniques of General-Purpose computing on Graphics Processing Units for large-scale parallel computing. Compared to the standard method in three-dimensional problems, our approach is more accurate and 133 times faster. This advantage increases when dimensionality increases. Our method is scalable to higher-dimensional problems since the precision bound is a finite number not affected by dimensionality. Our software will be available on GitHub and CRAN, which can be applied to guide the design of clinical trials to incorporate the biomarker better. Although our research is motivated by the design of clinical trials, the method can be used widely to solve other optimization problems involving high-dimensional integration., Comment: 18 pages, 3 figures, 1 table

Genome-wide association study (GWAS) is a popular approach to investigate relationships between genetic information and diseases. A number of associations are tested in a study and the results are often corrected using multiple adjustment methods. It is observed that GWAS studies suffer adequate statistical power for reliability. Hence, we document known models for reliability assessment using improved statistical power in GWAS analysis.

Summary Background The signaling protein p38 mitogen-activated protein kinase (MAPK) regulates the tumor cell microenvironment, modulating cell survival, migration, and invasion. This phase 1 study evaluated the safety of p38 MAPK inhibitor LY3007113 in patients with advanced cancer to establish a recommended phase 2 dose. Methods In part A (dose escalation), LY3007113 was administered orally every 12 h (Q12H) at doses ranging from 20 mg to 200 mg daily on a 28-day cycle until the maximum tolerated dose (MTD) was reached. In part B (dose confirmation), patients received MTD. Safety, pharmacokinetics, pharmacodynamics, and tumor response data were evaluated. Results MTD was 30 mg Q12H. The most frequent treatment-related adverse events (>10%) were tremor, rash, stomatitis, increased blood creatine phosphokinase, and fatigue. Grade ≥ 3 treatment-related adverse events included upper gastrointestinal haemorrhage and increased hepatic enzyme, both occurring at 40 mg Q12H and considered dose-limiting toxicities. LY3007113 exhibited an approximately dose-proportional increase in exposure and time-independent pharmacokinetics after repeated dosing. Maximal inhibition (80%) of primary biomarker MAPK-activated protein kinase 2 in peripheral blood mononuclear cells was not reached, and sustained minimal inhibition (60%) was not maintained for 6 h after dosing to achieve a biologically effective dose (BED). The best overall response in part B was stable disease in 3 of 27 patients. Conclusions The recommended phase 2 dosage of LY3007113 was 30 mg Q12H. Three patients continued treatment after the first radiographic assessment, and the BED was not achieved. Further clinical development of this compound is not planned as toxicity precluded achieving a biologically effective dose.

Motivation: HIV is difficult to treat because its virus mutates at a high rate and mutated viruses easily develop resistance to existing drugs. If the relationships between mutations and drug resistances can be determined from historical data, patients can be provided personalized treatment according to their own mutation information. The HIV Drug Resistance Database was built to investigate the relationships. Our goal is to build a model using data in this database, which simultaneously predicts the resistance of multiple drugs using mutation information from sequences of viruses for any new patient. Results: We propose two variations of a stacking algorithm which borrow information among multiple prediction tasks to improve multivariate prediction performance. The most attractive feature of our proposed methods is the flexibility with which complex multivariate prediction models can be constructed using any univariate prediction models. Using cross-validation studies, we show that our proposed methods outperform other popular multivariate prediction methods. Availability: An R package will be made available., Comment: 15 pages, 5 figures

Purpose: p38 MAPK regulates the production of cytokines in the tumor microenvironment and enables cancer cells to survive despite oncogenic stress, radiotherapy, chemotherapy, and targeted therapies. Ralimetinib (LY2228820 dimesylate) is a selective small-molecule inhibitor of p38 MAPK. This phase I study aimed to evaluate the safety and tolerability of ralimetinib, as a single agent and in combination with tamoxifen, when administered orally to patients with advanced cancer. Experimental Design: The study design consisted of a dose-escalation phase performed in a 3+3 design (Part A; n = 54), two dose-confirmation phases [Part B at 420 mg (n = 18) and Part C at 300 mg (n = 8)], and a tumor-specific expansion phase in combination with tamoxifen for women with hormone receptor–positive metastatic breast cancer refractory to aromatase inhibitors (Part D; n = 9). Ralimetinib was administered orally every 12 hours on days 1 to 14 of a 28-day cycle. Results: Eighty-nine patients received ralimetinib at 11 dose levels (10, 20, 40, 65, 90, 120, 160, 200, 300, 420, and 560 mg). Plasma exposure of ralimetinib (Cmax and AUC) increased in a dose-dependent manner. After a single dose, ralimetinib inhibited p38 MAPK–induced phosphorylation of MAPKAP-K2 in peripheral blood mononuclear cells. The most common adverse events, possibly drug-related, included rash, fatigue, nausea, constipation, pruritus, and vomiting. The recommended phase II dose was 300 mg every 12 hours as monotherapy or in combination with tamoxifen. Although no patients achieved a complete response or partial response,19 patients (21.3%) achieved stable disease with a median duration of 3.7 months, with 9 of these patients on study for ≥6 cycles. Conclusions: Ralimetinib demonstrated acceptable safety, tolerability, and pharmacokinetics for patients with advanced cancer. Clin Cancer Res; 22(5); 1095–102. ©2015 AACR.

Glioblastoma is a type of glioma with a relatively higher degree of malignancy that may result in severe intracranial hypertension and focal symptoms. Surgery is the preferred treatment modality. Combination therapy including radiotherapy, chemotherapy, gene therapy, immunotherapy and targeted therapy have also been employed. However, due to the invasiveness and pathogenesis of the disease, such treatments do not yield satisfactory outcomes. The aim of the present study was to examine the expression of microRNA (miR)-184 in Janus kinase 2/signal transducer and activator of transcription 3 (JAK2/STAT3) signaling pathway in the mechanism of glioblastoma formation, thus providing a new basis for the mechanism of glioblastoma induction. The LN18 cell line was employed in the present study. After undergoing thawing, culturing and passaging processes, the cells were divided into the set control group, miR-184 mimic group (transfer miR-184 simulator) and miR-184 group. The expression of miR-184 was detected using quantitative polymerase chain reaction. An MTT assay was used to detect the proliferation ability of glioma cells, and clone formation ability was also detected. The cell scratch and invasion assays were used to identify the cell invasion ability. Western blotting was performed to detect the expression level of p-JAK2 and p-STAT3 proteins. The results showed that compared to the control group, the expression of miR-184 in the miR-184 mimic group increased. Cell proliferation, as well as clone formation and invasion ability were enhanced. The number of cells penetrating septum, as well as the expression of p-JAK2 and p-STAT3 proteins were increased. Differences were statistically significant (P

the vascular endothelial growth factor (VEGF) pathway plays a prominent role in the growth and progression of human cancer, including non-small cell lung carcinoma (NSCLC). The key mediators of VEGF signaling are a family of related receptor tyrosine kinases that include VEGFR1, VEGFR2, and VEGFR3. The relative expression levels, activity, and cross-talk among these receptors may contribute to response of NSCLC to anti-angiogenic therapies, and a better systematic, translatable approach to categorizing tumors is needed.We comparatively evaluated immunohistochemical expression of the three VEGFRs in archival primary NSCLC tissues (n=96).VEGFR1 and VEGFR2 were localized both in vessels and tumor cells, while VEGFR3 was only localized in tumor vessels. A set of eight VEGFR staining subclasses were identified: Triple VEGFR positive (n=11, 11.5%), VEGFR1 predominant (n=22, 22.9%), VEGFR2 predominant (n=9, 9.4%), VEGFR3 predominant (n=3, 3.1%), VEGFR1/2 predominant (13, 13.5%), VEGFR1/3 predominant (2, 2.1%), VEGFR2/3 predominant (n=8, 8.3%), and triple VEGFR negative (n=28, 29.2%). An objective categorization based on K-means clustering revealed four clusters, three of which showed high VEGFR2 compared to VEGFR3 (30.7% of cases), cases high in both VEGFR2 and VEGFR3 (18.2%), and cases that were negative/low for both VEGFR2 and VEGFR3 (45.4%). A positive association between VEGFR2 and VEGFR3 was found, however no associations were observed between VEGFR1 and VEGFR2, nor VEGFR1 and VEGFR3.The proposed subclasses of NSCLC are an approach for complementing lines of investigation of anti-angiogenic therapies beginning with systematic characterization of the disease.

Summary Background Concern is continuing about increased risk of pneumonia in patients with chronic obstructive pulmonary disease (COPD) who use inhaled corticosteroids. We aimed to establish the effects of inhaled budesonide on the risk of pneumonia in such patients. Methods We pooled patient data from seven large clinical trials of inhaled budesonide (320–1280 μg/day), with or without formoterol, versus control regimen (placebo or formoterol alone) in patients with stable COPD and at least 6 months of follow-up. The primary analysis compared treatment groups for the risk of pneumonia as an adverse event or serious adverse event during the trial or within 15 days of the trial end. Cox proportional hazards regression was used to analyse the data on an intention-to-treat basis. Data were adjusted for patients' age, sex, smoking status, body-mass index, and postbronchodilator percent of predicted forced expiratory volume in 1 s (FEV 1 ). Findings We analysed data from 7042 patients, of whom 3801 were on inhaled budesonide and 3241 were on control treatment, with 5212 patient-years of exposure to treatment. We recorded no significant difference between treatment groups for the occurrence of pneumonia as an adverse event (3% [n=122 patients] vs 3% [n=103]; adjusted hazard ratio 1·05, 95% CI 0·81–1·37) or a serious adverse event (1% [n=53] vs 2% [n=50]; 0·92, 0·62–1·35), or for time to pneumonia as an adverse event (log-rank test 0·94) or a serious adverse event (0·61). Increasing age and decreasing percent of predicted FEV 1 were the only two variables that were significantly associated with occurrence of pneumonia as an adverse event or a serious adverse event. Interpretation Budesonide treatment for 12 months does not increase the risk of pneumonia in patients with COPD during that time and therefore is safe for clinical use in such patients. Funding Michael Smith Foundation for Health Research.

Background: Interleukin-6 (IL6) is a pleiotropic pro-inflammatory and immunomodulatory cytokine which probably plays an important role in the pathogenesis of chronic obstructive pulmonary disease (COPD). There is a functional single nucleotide polymorphism (SNP), -174G/C, in the promoter region of IL6 . It was hypothesised that IL6 SNPs influence susceptibility for impaired lung function and COPD in smokers. Methods: Seven and five SNPs in IL6 were genotyped in two nested case-control samples derived from the Lung Health Study (LHS) based on phenotypes of rate of decline of forced expiratory volume in 1 s (FEV1) over 5 years and baseline FEV1 at the beginning of the LHS. Serum IL6 concentrations were measured for all subjects. A partially overlapping panel of nine IL6 SNPs was genotyped in 389 cases of COPD from the National Emphysema Treatment Trial (NETT) and 420 controls from the Normative Aging Study (NAS). Results: In the LHS, three IL6 SNPs were associated with decline in FEV1 (0.023⩽p⩽0.041 in additive models). Among them, the IL6 \_-174C allele was associated with a rapid decline in lung function. The association was more significant in a genotype-based analysis (p = 0.006). In the NETT-NAS study, IL6 \_-174G/C and four other IL6 SNPs, all of which are in linkage disequilibrium with IL6 _-174G/C, were associated with susceptibility to COPD (0.01⩽p⩽0.04 in additive genetic models). Conclusion: The results suggest that the IL6 _-174G/C SNP is associated with a rapid decline in FEV1 and susceptibility to COPD in smokers.

Interleukin (IL)-10 is a type-2 T-helper cell cytokine with a broad spectrum of anti-inflammatory actions. Inflammation plays an important role in the pathogenesis of chronic obstructive pulmonary disease. It was hypothesised that single nucleotide polymorphisms (SNPs) of the genes encoding IL-10 ( IL10 ) and the α subunit of its receptor ( IL10RA ) are associated with changes in, or value of, forced expiratory volume in one second (FEV 1 ) in smoking-induced chronic obstructive pulmonary disease. In total, eleven SNPs of IL10 and IL10RA were studied in 586 White subjects, selected from continuous smokers followed for 5 yrs in the Lung Health Study, who showed the fastest (n = 280) and slowest (n = 306) decline in FEV 1 . These 11 SNPs were also studied in 1,072 participants exhibiting the lowest (n = 538) and highest (n = 534) baseline FEV 1 at the beginning of the Lung Health Study. No association was found in the primary analyses. Although a subgroup analysis showed that the IL-10 3368A allele was associated with a fast decline in FEV 1 , the association did not pass correction for multiple comparisons. No gene–gene interaction of IL10 with IL10RA was found. There was no association of polymorphisms of the genes encoding interleukin-10 and the α subunit of its receptor with the rate of decline in, or value of, forced expiratory volume in one second in smoking-induced chronic obstructive pulmonary disease.

Background: Galunisertib, a selective transforming growth factor-β receptor1 inhibitor, is being investigated in clinical trials for hepatocellular carcinoma (HCC). MicroRNAs (miRs) are small (∼22 nucleotide) non-coding RNAs that regulate expression of targeted genes, and are secreted by cells into blood. miRs are differentially expressed during HCC progression, differentiation, and response to therapy. We hypothesize that circulating miRs may be useful to identify patients who benefit from galunisertib treatment. Patients and Methods: Plasma samples from HCC patients (n = 105) treated with galunisertib were analyzed for miR expression. Patients were enrolled in part A, of a multi-part single-arm study in 2nd-line HCC (phase II trial NCT01246986/JBAK). All patients had elevated alpha-fetoprotein (AFP) levels at baseline (AFP ≥1.5 x ULN). The median OS of this cohort of patients was 7.2 mo. Eighty percent of patients received prior sorafenib treatment. Plasma samples were collected during patient screening, cycle 1 day 1 (pretreatment), and cycle 2 day 14. The Exiqon RT microRNA PCR Human panel I+II was used to measure 752 miRs. Expression levels of detectable miRs and their association with overall survival (OS) were investigated. Results: Low plasma levels of miR-665 (HR = 0.50, p = 0.001, q = 0.13), miR-320d (HR = 0.44, p = 0.001, q = 0.13), miR-320a (HR = 0.47, p = 0.001, q = 0.13), and miR-130b-3p (HR = 0.44, p = 0.002, q = 0.13) are associated with better survival. Whereas, low plasma levels of miR-451a (HR = 2.0, p = 0.002, q = 0.13), miR-let7g-5p (HR = 2.3, p = 0.002, q = 0.13), and miR-18a-3p (HR = 2.0, p = 0.002, q = 0.13) are associated with poor survival. To assess within patient baseline biological variation of miR expression, a comparison of the expression of 369 miRs in patients (n = 42) with 2 pre-treatment samples was performed. The proportion of miRs attaining statistical significance was smaller than what we would expect by chance. Conclusions: Circulating miRs may serve as easily accessible markers to identify HCC patients who may benefit from galunisertib treatment, which requires confirmation in randomized controlled study. Given the low intra-patient variability measured at baseline for most of the miRs, circulating miRs may represent reliable molecular markers with prognostic and/or predictive utility. Citation Format: Shawn T. Estrem, Michael Man, Xuekui Zhang, Duytrac Nguyen, Danni Yu, Michael M. Lahn, Ann Cleverly, Durisala Desaiah, Sandrine Faivre, Gianluigi Gianneli, Karim A. Benhadji. Plasma microRNAs associated with overall survival in patients with hepatocellular carcinoma treated with galunisertib (LY2157299 monohydrate), an inhibitor of transforming growth factor-β receptor1. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 1076.

The paucity of information on the epigenetic barriers that are blocking reprogramming protocols, and on what makes a beta cell unique, has hampered efforts to develop novel beta cell sources. Here, we aimed to identify enhancers in pancreatic islets, to understand their developmental ontologies, and to identify enhancers unique to islets to increase our understanding of islet-specific gene expression.We combined H3K4me1-based nucleosome predictions with pancreatic and duodenal homeobox 1 (PDX1), neurogenic differentiation 1 (NEUROD1), v-Maf musculoaponeurotic fibrosarcoma oncogene family, protein A (MAFA) and forkhead box A2 (FOXA2) occupancy data to identify enhancers in mouse islets.We identified 22,223 putative enhancer loci in in vivo mouse islets. Our validation experiments suggest that nearly half of these loci are active in regulating islet gene expression, with the remaining regions probably poised for activity. We showed that these loci have at least nine developmental ontologies, and that islet enhancers predominately acquire H3K4me1 during differentiation. We next discriminated 1,799 enhancers unique to islets and showed that these islet-specific enhancers have reduced association with annotated genes, and identified a subset that are instead associated with novel islet-specific long non-coding RNAs (lncRNAs).Our results indicate that genes with islet-specific expression and function tend to have enhancers devoid of histone methylation marks or, less often, that are bivalent or repressed, in embryonic stem cells and liver. Further, we identify a subset of enhancers unique to islets that are associated with novel islet-specific genes and lncRNAs. We anticipate that these data will facilitate the development of novel sources of functional beta cell mass.

We describe a model-based method, PING, for predicting nucleosome positions in MNase-Seq and MNase- or sonicated-ChIP-Seq data. PING compares favorably to NPS and TemplateFilter in scalability, accuracy and robustness to low read density. To demonstrate that PING predictions from widely available sonicated data can have sufficient spatial resolution to be to be useful for biological inference, we use Illumina H3K4me1 ChIP-seq data to detect changes in nucleosome positioning around transcription factor binding sites due to tamoxifen stimulation, to discriminate functional and non-functional transcription factor binding sites more effectively than with enrichment profiles, and to confirm that the pioneer transcription factor Foxa2 associates with the accessible major groove of nucleosomal DNA.

Rationale: There are no accepted blood-based biomarkers in chronic obstructive pulmonary disease (COPD). Pulmonary and activation-regulated chemokine (PARC/CCL-18) is a lung-predominant inflammatory protein that is found in serum. Objectives: To determine whether PARC/CCL-18 levels are elevated and modifiable in COPD and to determine their relationship to clinical end points of hospitalization and mortality. Methods: PARC/CCL-18 was measured in serum samples from individuals who participated in the ECLIPSE (Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints) and LHS (Lung Health Study) studies and a prednisolone intervention study. Measurements and Main Results: Serum PARC/CCL-18 levels were higher in subjects with COPD than in smokers or lifetime nonsmokers without COPD (105 vs. 81 vs. 80 ng/ml, respectively; P < 0.0001). Elevated PARC/CCL-18 levels were associated with increased risk of cardiovascular hospitalization or mortality in the LHS cohort and with total mortality in the ECLIPSE cohort. Conclusions: Serum PARC/CCL-18 levels are elevated in COPD and track clinical outcomes. PARC/CCL-18, a lung-predominant chemokine, could be a useful blood biomarker in COPD. Clinical trial registered with www.clinicaltrials.gov (NCT 00292552).

ChIP-seq, which combines chromatin immunoprecipitation with massively parallel short-read sequencing, can profile in vivo genome-wide transcription factor-DNA association with higher sensitivity, specificity and spatial resolution than ChIP-chip. While it presents new opportunities for research, ChIP-seq poses new challenges for statistical analysis that derive from the complexity of the biological systems characterized and the variability and biases in its digital sequence data. We propose a method called PICS (Probabilistic Inference for ChIP-seq) for extracting information from ChIP-seq aligned-read data in order to identify regions bound by transcription factors. PICS identifies enriched regions by modeling local concentrations of directional reads, and uses DNA fragment length prior information to discriminate closely adjacent binding events via a Bayesian hierarchical t-mixture model. Its per-event fragment length estimates also allow it to remove from analysis regions that have atypical lengths. PICS uses pre-calculated, whole-genome read mappability profiles and a truncated tdistribution to adjust binding event models for reads that are missing due to local genome repetitiveness. It estimates uncertainties in model parameters that can be used to define confidence regions on binding event locations and to filter estimates. Finally, PICS calculates a per-event enrichment score relative to a control sample, and can use a control sample to estimate a false discovery rate. We compared PICS to the alternative methods MACS, QuEST, and CisGenome, using published GABP and FOXA1 data sets from human cell lines, and found that PICS’ predicted binding sites were more consistent with computationally predicted binding motifs.

Background: Several studies suggest that inhaled and oral corticosteroids repress systemic inflammation in chronic obstructive pulmonary disease (COPD). However, the cytokines that may respond to these medications are unclear. Method: We used data from 41 patients with a history of stable moderate COPD (average age 64 years) who were randomised to inhaled fluticasone (500 μg twice daily from a Diskus inhaler), oral prednisone (30 mg daily) or placebo for 2 weeks. Using a multiplexed array system, different serum cytokines that have been implicated in COPD pathogenesis were measured. Results: We found that compared with placebo, inhaled fluticasone significantly reduced levels of soluble tumour necrosis factor receptor-2 (sTNF-R2) by 24% (95% CI, 7—38%; p = 0.01), monocyte chemoattractant protein-1 by 20% (95% CI, 5—32%; p = 0.01), interferon gamma inducible CXCL10 (IP-10) by 43% (95% CI, 3—66%; p = 0.04), and soluble L-selectin levels by 15% (95% CI, 1—28%; p = 0.04). Compared with placebo, oral prednisone reduced levels of sTNF-R2 by 26% (95% CI, 15—36%; p < 0.001), L-selectin by 22% (95% CI, 8—34%; p = 0.004), intercellular adhesion molecule-1 by 31% (95% CI, 9—48%; p = 0.01), pulmonary and activation-regulated chemokine (PARC) by 18% (95% CI, 2—32%; p = 0.03) and IP-10 by 40% (95% CI, 0—64%; p = 0.05). sTNF-R2, L-selectin and IP-10 were significantly reduced by both oral and inhaled corticosteroids. The other cytokines were not significantly repressed by either oral or inhaled corticosteroids. Conclusions: In summary, inhaled and oral corticosteroids significantly repressed a selected number of systemic cytokines in patients with stable, moderate COPD; most of the steroid-responsive cytokines appear to be chemoattractants.

The balance between inflammatory and repair processes is important in maintaining lung homeostasis in chronic obstructive pulmonary disease (COPD). The aim of the present study was to determine whether or not an integrated index of a biomarker involved in inflammation, C-reactive protein (CRP), and another involved in wound repair, fibronectin, may be a good measure to predict clinical outcomes in COPD. Circulating blood levels of CRP and fibronectin were measured in 4,787 individuals with mild-to-moderate COPD who were prospectively followed for7 yrs after blood collection as part of the Lung Health Study. To assess the balance between repair and inflammation, a simple ratio was calculated by dividing fibronectin levels by CRP levels and a Cox proportional hazards model was used to determine the relationship between this ratio and all-cause and disease-specific causes of mortality. The relationship between the fibronectin to CRP ratio and all-cause mortality was L-shaped. There was an exponential decay in the adjusted hazard function (i.e. the risk of mortality) as the ratio decreased until a value of 148 was reached, beyond which point the hazard function did not change significantly. Similar results were observed for the risk of coronary and cardiovascular mortality. Circulating fibronectin to CRP ratio is significantly associated with all-cause mortality of COPD patients. However, in contrast to other biomarkers, the relationship appears to be L-shaped (and not linear), suggesting a threshold at approximately 150. While promising, future studies are needed to validate this simple index as a biomarker in COPD.

Epidemiologic and animal studies have shown that exposure to particulate matter air pollution (PM) is a risk factor for the development of atherosclerosis. Whether PM-induced lung and systemic inflammation is involved in this process is not clear. We hypothesized that PM exposure causes lung and systemic inflammation, which in turn leads to vascular endothelial dysfunction, a key step in the initiation and progression of atherosclerosis. New Zealand White rabbits were exposed for 5 days (acute, total dose 8 mg) and 4 wk (chronic, total dose 16 mg) to either PM smaller than 10 μm (PM10) or saline intratracheally. Lung inflammation was quantified by morphometry; systemic inflammation was assessed by white blood cell and platelet counts and serum interleukin (IL)-6, nitric oxide, and endothelin levels. Endothelial dysfunction was assessed by vascular response to acetylcholine (ACh) and sodium nitroprusside (SNP). PM10 exposure increased lung macrophages ( P < 0.02), macrophages containing particles ( P < 0.001), and activated macrophages ( P < 0.006). PM10 increased serum IL-6 levels in the first 2 wk of exposure ( P < 0.05) but not in weeks 3 or 4. PM10 exposure reduced ACh-related relaxation of the carotid artery with both acute and chronic exposure, with no effect on SNP-induced vasodilatation. Serum IL-6 levels correlated with macrophages containing particles ( P = 0.043) and ACh-induced vasodilatation ( P = 0.014 at week 1, P = 0.021 at week 2). Exposure to PM10 caused lung and systemic inflammation that were both associated with vascular endothelial dysfunction. This suggests that PM-induced lung and systemic inflammatory responses contribute to the adverse vascular events associated with exposure to air pollution.

Background: There is controversy about whether therapy with inhaled corticosteroids (ICSs) modifies the natural history of COPD, characterized by an accelerated decline in FEV1.Methods: The Inhaled Steroids Effect Evaluation in COPD (ISEEC) study is a pooled study of patient-level data from seven long-term randomized controlled trials of ICS vs placebo lasting >= 12 months in patients with moderate-to-severe COPD. We have previously reported a survival benefit for ICS therapy in COPD patients using ISEEC data. We aimed to determine whether the regular use of ICSs vs placebo improves FEV1 decline in COPD patients, and whether this relationship is modified by gender and smoking.Results: There were 3,911 randomized participants (29.2% female) in this analysis. In the first 6 months after randomization, ICS use was associated with a significant mean (+/- SE) relative increase in FEV1 of 2.42 +/- 0.19% compared with placebo (p Conclusions: We conclude that in COPD in the first 6 months of treatment, ICS therapy is more effective in ex-smokers than in current smokers with COPD in improving lung function, and women may have a bigger response to ICSs than men. However, it seems that after 6 months, ICS therapy does not modify the decline in FEV1 among those who completed these randomized clinical trials.

Summary: MNase-Seq and ChIP-Seq have evolved as popular techniques to study chromatin and histone modification. Although many tools have been developed to identify enriched regions, software tools for nucleosome positioning are still limited. We introduce a flexible and powerful open-source R package, PING 2.0, for nucleosome positioning using MNase-Seq data or MNase– or sonicated– ChIP-Seq data combined with either single-end or paired-end sequencing. PING uses a model-based approach, which enables nucleosome predictions even in the presence of low read counts. We illustrate PING using two paired-end datasets from Saccharomyces cerevisiae and compare its performance with nucleR and ChIPseqR. Availability: PING 2.0 is available from the Bioconductor website at http://bioconductor.org. It can run on Linux, Mac and Windows. Contact: rgottard@fhcrc.org Supplementary Information: Supplementary material is available at Bioinformatics online.