Peter W. Li

Multi-Platform Analysis of MicroRNA Expression Measurements in RNA from Fresh Frozen and FFPE Tissues

MicroRNAs play a role in regulating diverse biological processes and have considerable utility as molecular markers for diagnosis and monitoring of human disease. Several technologies are available commercially for measuring microRNA expression. However, cross-platform comparisons do not necessarily correlate well, making it difficult to determine which platform most closely represents the true microRNA expression level in a tissue. To address this issue, we have analyzed RNA derived from cell lines, as well as fresh frozen and formalin-fixed paraffin embedded tissues, using Affymetrix, Agilent, and Illumina microRNA arrays, NanoString counting, and Illumina Next Generation Sequencing. We compared the performance within- and between the different platforms, and then verified these results with those of quantitative PCR data. Our results demonstrate that the within-platform reproducibility for each method is consistently high and although the gene expression profiles from each platform show unique traits, comparison of genes that were commonly detectable showed that detection of microRNA transcripts was similar across multiple platforms.

Common Oncogene Mutations and Novel SND1-BRAF Transcript Fusion in Lung Adenocarcinoma from Never Smokers

Lung adenocarcinomas from never smokers account for approximately 15 to 20% of all lung cancers and these tumors often carry genetic alterations that are responsive to targeted therapy. Here we examined mutation status in 10 oncogenes among 89 lung adenocarcinomas from never smokers. We also screened for oncogene fusion transcripts in 20 of the 89 tumors by RNA-Seq. In total, 62 tumors had mutations in at least one of the 10 oncogenes, including EGFR (49 cases, 55%), K-ras (5 cases, 6%), BRAF (4 cases, 5%), PIK3CA (3 cases, 3%), and ERBB2 (4 cases, 5%). In addition to ALK fusions identified by IHC/FISH in four cases, two previously known fusions involving EZR- ROS1 and KIF5B-RET were identified by RNA-Seq as well as a third novel fusion transcript that was formed between exons 1–9 of SND1 and exons 2 to 3′ end of BRAF. This in-frame fusion was observed in 3/89 tested tumors and 2/64 additional never smoker lung adenocarcinoma samples. Ectopic expression of SND1-BRAF in H1299 cells increased phosphorylation levels of MEK/ERK, cell proliferation, and spheroid formation compared to parental mock-transfected control. Jointly, our results suggest a potential role of the novel BRAF fusion in lung cancer development and therapy.

Inflammation and the depot-specific secretome of human preadipocytes

Visceral white adipose tissue (WAT) expansion and macrophage accumulation are associated with metabolic dysfunction. Visceral WAT typically shows greater macrophage infiltration. Preadipocytes show varying proinflammatory expression profiles among WAT depots. The objective was to examine the secretomes and chemoattractive properties of preadipocytes from visceral and subcutaneous WAT.

A simple statistical model and association rule filtering for classification

Associative classification is a predictive modeling technique that constructs a classifier based on class association rules (also known as predictive association rules; PARs). PARs are association rules where the consequence of the rule is a class label. Associative classification has gained substantial research attention because it successfully joins the benefits of association rule mining with classification. These benefits include the inherent ability of association rule mining to extract high-order interactions among the predictors--an ability that many modern classifiers lack--and also the natural interpretability of the individual PARs. Associative classification is not without its caveats. Association rule mining often discovers a combinatorially large number of association rules, eroding the interpretability of the rule set. Extensive effort has been directed towards developing interestingness measures, which filter (predictive) association rules after they have been generated. These interestingness measures, albeit very successful at selecting interesting rules, lack two features that are highly valuable in the context of classification. First, only few of the interestingness measures are rooted in a statistical model. Given the distinction between a training and a test data set in the classification setting, the ability to make statistical inferences about the performance of the predictive classification rules on the test set is highly desirable. Second, the unfiltered set of predictive assocation rules (PARs) are often redundant, we can prove that certain PARs will not be used to construct a classification model given the presence of other PARs. In this paper, we propose a simple statistical model towards making inferences on the test set about the various performance metrics of predictive association rules. We also derive three filtering criteria based on hypothesis testing, which are very selective (reduce the number of PARs to be considered by the classifier by several orders of magnitude), yet do not effect the performance of the classification adversely. In the case, where the classification model is constructed as a logistic model on top of the PARs, we can mathematically prove, that the filtering criteria do not significantly effect the classifiers performance. We also demonstrate empirically on three publicly available data sets that the vast reduction in the number of PARs indeed did not come at the cost of reducing the predictive performance.

Extending Association Rule Summarization Techniques to Assess Risk of Diabetes Mellitus

Early detection of patients with elevated risk of developing diabetes mellitus is critical to the improved prevention and overall clinical management of these patients. We aim to apply association rule mining to electronic medical records (EMR) to discover sets of risk factors and their corresponding subpopulations that represent patients at particularly high risk of developing diabetes. Given the high dimensionality of EMRs, association rule mining generates a very large set of rules which we need to summarize for easy clinical use. We reviewed four association rule set summarization techniques and conducted a comparative evaluation to provide guidance regarding their applicability, strengths and weaknesses. We proposed extensions to incorporate risk of diabetes into the process of finding an optimal summary. We evaluated these modified techniques on a real-world prediabetic patient cohort. We found that all four methods produced summaries that described subpopulations at high risk of diabetes with each method having its clear strength. For our purpose, our extension to the Buttom-Up Summarization (BUS) algorithm produced the most suitable summary. The subpopulations identified by this summary covered most high-risk patients, had low overlap and were at very high risk of diabetes.

ReliefSeq: a gene-wise adaptive-K nearest-neighbor feature selection tool for finding gene-gene interactions and main effects in mRNA-Seq gene expression data.

Relief-F is a nonparametric, nearest-neighbor machine learning method that has been successfully used to identify relevant variables that may interact in complex multivariate models to explain phenotypic variation. While several tools have been developed for assessing differential expression in sequence-based transcriptomics, the detection of statistical interactions between transcripts has received less attention in the area of RNA-seq analysis. We describe a new extension and assessment of Relief-F for feature selection in RNA-seq data. The ReliefSeq implementation adapts the number of nearest neighbors (k) for each gene to optimize the Relief-F test statistics (importance scores) for finding both main effects and interactions. We compare this gene-wise adaptive-k (gwak) Relief-F method with standard RNA-seq feature selection tools, such as DESeq and edgeR, and with the popular machine learning method Random Forests. We demonstrate performance on a panel of simulated data that have a range of distributional properties reflected in real mRNA-seq data including multiple transcripts with varying sizes of main effects and interaction effects. For simulated main effects, gwak-Relief-F feature selection performs comparably to standard tools DESeq and edgeR for ranking relevant transcripts. For gene-gene interactions, gwak-Relief-F outperforms all comparison methods at ranking relevant genes in all but the highest fold change/highest signal situations where it performs similarly. The gwak-Relief-F algorithm outperforms Random Forests for detecting relevant genes in all simulation experiments. In addition, Relief-F is comparable to the other methods based on computational time. We also apply ReliefSeq to an RNA-Seq study of smallpox vaccine to identify gene expression changes between vaccinia virus-stimulated and unstimulated samples. ReliefSeq is an attractive tool for inclusion in the suite of tools used for analysis of mRNA-Seq data; it has power to detect both main effects and interaction effects. Software Availability: http://insilico.utulsa.edu/ReliefSeq.php.

Development of a Semi-synthetic Dataset as a Testbed for Big-Data Semantic Analytics

We have developed a large semi-synthetic, semantically rich dataset, modeled after the medical record of a large medical institution. Using the highly diverse data.gov data repository and a multivariate data augmentation strategy, we can generate arbitrarily large semi-synthetic datasets which can be used to test new algorithms and computational platforms. The construction process and basic data characterization are described. The databases, as well as code for data collection, consolidation, and augmentation are available for distribution.

Generation and Sequencing of Pulmonary Carcinoid Tumor Cell Lines

Introduction: Pulmonary carcinoid tumors account for approximately 5% of all lung malignancies in adults, and comprise 30% of all carcinoid tumors. There are limited reagents available to study these rare tumors, and consequently no major advances have been made for patient treatment. We report the generation and characterization of human pulmonary carcinoid tumor cell lines to study underlying biology, and to provide models for testing novel chemotherapeutic agents. Methods: Tissue was harvested from three patients with primary pulmonary typical carcinoid tumors undergoing surgical resection. The tumor was dissociated and plated onto dishes in culture media. The established cell lines were characterized by immunohistochemistry, Western blotting, and cell proliferation assays. Tumorigenicity was confirmed by soft agar growth and the ability to form tumors in a mouse xenograft model. Exome and RNA sequencing of patient tumor samples and cell lines was performed using standard protocols. Results: Three typical carcinoid tumor lines grew as adherent monolayers in vitro, expressed neuroendocrine markers consistent with the primary tumor, and formed colonies in soft agar. A single cell line produced lung tumors in nude mice after intravenous injection. Exome and RNA sequencing of this cell line showed lineage relationship with the primary tumor, and demonstrated mutations in a number of genes related to neuronal differentiation. Conclusion: Three human pulmonary typical carcinoid tumor cell lines have been generated and characterized as a tool for studying the biology and novel treatment approaches for these rare tumors.

Understanding atrophy trajectories in alzheimer's disease using association rules on MRI images

Alzheimers disease (AD) is associated with progressive cognitive decline leading to dementia. The atrophy/loss of brain structure as seen on Magnetic Resonance Imaging (MRI) is strongly correlated with the severity of the cognitive impairment in AD. In this paper, we set out to find associations between predefined regions of the brain (regions of interest; ROIs) and the severity of the disease. Specifically, we use these associations to address two important issues in AD: (i) typical versus atypical atrophy patterns and (ii) the origin and direction of progression of atrophy, which is currently under debate. We observed that each AD-related ROI is associated with a wide range of severity and that the difference between ROIs is merely a difference in severity distribution. To model differences between the severity distribution of a subpopulation (with significant atrophy in certain ROIs) and the severity distribution of the entire population, we developed the concept of Distributional Association Rules. Using the Distributional Association Rules, we clustered ROIs into disease subsystems. We define a disease subsystem as a contiguous set of ROIs that are collectively implicated in AD. AD is known to be heterogeneous in the sense that multiple sets of ROIs may be related to the disease in a population. We proposed an enhancement to the association rule mining where the algorithm only discovers association rules with ROIs that form an approximately contiguous volume. Next, we applied these association rules to infer the direction of disease progression based on the support measures of the association rules. We also developed a novel statistical test to determine the statistical significance of the discovered direction. We evaluated the proposed method on the Mayo Clinic Alzheimers Disease Research Center (ADRC) prospective patient cohorts. The key achievements of the methodology is that it accurately identified larger disease subsystems implicated in typical and atypical AD and it successfully mapped the directions of disease progression. The wealth of data available in Radiology gives rise to opportunities for applying this methodology to map out the trajectory of several other diseases, e.g. other neuro-degenerative diseases and cancers, most notably, breast cancer. The applicability of this method is not limited to image data, as associating predictors with severity provides valuable information in most areas of medicine as well as other industries.

Characterization of semi-synthetic dataset for big-data semantic analysis

Over the past decade, the use of semantic databases has served as the basis for storing and analyzing complex, heterogeneous, and irregular data. While there are similarities with traditional relational database systems, semantic data stores provide a rich platform for conducting non-traditional analyses of data. In support of new graph analytic algorithms and specialized graph analytic hardware, we have developed a large semi-synthetic, semantically rich dataset. The construction of this dataset mimics the real-world scenario of using relational databases as the basis for semantic data construction. In order to achieve real-world variable distributions and variable dependencies, data.gov data was used as the basis for developing an approach to build arbitrarily large semi-synthetic datasets. The intent of the semi-synthetic dataset is to serve as a testbed for new semantic graph analyses and computational software/hardware platforms. The construction process and basic data characterization is described. All code related to the data collection, consolidation, and augmentation are available for distribution.