No description available.

Cartas al editor

Genome-wide RNAi screening has emerged as a powerful tool for loss-of-function studies that may lead to therapeutic target discovery for human malignancies in the era of personalized medicine. However, due to high false-positive and false-negative rates arising from noise of high-throughput measurements and off-target effects, powerful computational tools and additional knowledge are much needed to analyze and complement it. Availability of high-throughput genomic data including gene expression profiles, copy number variations from large-sampled primary patients and cell lines allows us to tackle underlying drivers causally associated with tumorigenesis or drug-resistance. In my dissertation, I have developed a framework to integrate functional RNAi screens with systems biology of cancer genomics to tailor potential therapeutics for reversal of drug-resistance or treatment of aggressive tumors. I developed a series of algorithms and tools to deconvolute, QC and post-analyze high-throughput shRNA screening data by next-generation sequencing technology (shSeq), particularly a novel Bayesian hierarchical modeling approach to integrate multiple shRNAs targeting the same gene, which outperforms existing methods. In parallel, I developed a systems biology algorithm, NetBID2, to infer disease drivers from high-throughput genomic data by reverse-engineering network and Bayesian inference, which is able to detect hidden drivers that traditional methods fail to find. Integrating NetBID2 with functional RNAi screens, I have identified known and novel driver-type therapeutic targets in various disease contexts. For example, I discovered that AKT1 is a driver for glucocorticoid (GC) resistance, a problem in the treatment of T-ALL. The inhibition of AKT1 was validated to reverse GC-resistance. Additionally, upon silencing predicted master regulators of GC resistance with shRNA screens, 13 out of 16 were validated to significantly overcome resistance. In breast cancer, I discovered that STAT3 is required for transformation of HER2+ breast cancer, an aggressive breast tumor subtype. The suppression of STAT3 was confirmed in vitro and in vivo to be an effective therapy for HER2+ breast cancer. Moreover, my analysis revealed that STAT3 silencing only works in ER- cases. Using my framework, I have also identified potential therapeutic targets for ABC or GCB-type DLBCL and subtype-based breast cancer that are currently being validated.

Cancer research, like many areas of science, is adapting to a new era characterized by increasing quantity, quality, and diversity of observational data. An example of the advances, and the resulting challenges, is represented by The Cancer Genome Atlas, an enormous public effort that has provided genomic profiles of hundreds of tumors of each of the most common solid cancer types. Alongside this resource is a host of other data and knowledge, including gene interaction databases, Mendelian disease causal variants, and electronic health records spanning many millions of patients. Thus, a current challenge is how best to integrate these data to discover mechanisms of oncogenesis and cancer progression. Ultimately, this could enable genomics-based prediction of an individual patient's outcome and targeted therapies, a goal termed precision medicine. In this thesis, I develop novel approaches that examine patterns in populations of cancer patients to identify key genetic changes and suggest likely roles of these driver genes in the diseases. In the first section I show how genomics can lead to the identification of driver alterations in melanoma. The most recurrent genetic mutations are often in important cancer driver genes: in a newly sequenced melanoma cohort, recurrent inactivating mutations point to an exciting new melanoma candidate tumor suppressor, FBXW7, with therapeutic implications. But each tumor is unique, underlining the fact that recurrence will never capture all relevant mutations responsible for the disease. Tumors are a result of random events that must collaborate to endow a cell with all of the invasive and immortal properties of a cancer. Some combinations of events are lethal to a developing tumor, while other combinations are simply not preferentially selected. In order to discover these complex patterns, I develop a method based on the joint entropy of a set of genes, called GAMToC. Using GAMToC, I identify sets of recurrently altered genes with a strongly non-random joint pattern of co-occurrence and mutual exclusivity. Then, I extend this method as a means of identifying novel genes with a role in cancer, by virtue of their non-random pattern of alteration. Insights into the roles of these novel drivers can come from their most strongly co-selected partners. In the final section of the main text, I develop the use of cancer comorbidity, or increased cancer risk, as a novel data source for understanding cancer. The recent availability of clinical records spanning a large percentage of the American population has enabled discovery of many cancer comorbidities. Although most cancers arise as a result of somatic mutations accumulating over a patient's lifespan, mutations present at birth could predispose some rare populations to increased cancer risk. Mendelian disease phenotype provides strong insight into the genotype of an afflicted individual. Thus, if Mendelian diseases with cancer comorbidity can be shown to have specific defects in processes that are important in the development of that cancer, statistical comorbidity could provide a new a resource for prioritizing Mendelian disease genes as novel cancer related genes. For this purpose, I integrate clinical comorbidity, Mendelian disease causal variants, and somatic genomic profiles of thousands of cancers. I demonstrate that comorbidity indeed is associated with significant genetic similarity between Mendelian diseases and the cancers these patients are predisposed to, suggesting highly interesting and plausible new candidate cancer genes. While cancer may be the result of a series of selected random events, patterns of incidence across large populations, as measured by genomics or by other phenotypes, contain much non-random signal yet to be mined.

Ko Hon Man. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2005. / Includes bibliographical references (leaves 68-72). / Abstracts in English and Chinese. / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Background and Motivation --- p.1 / Chapter 1.2 --- Thesis Organization --- p.4 / Chapter 2 --- Related Work --- p.6 / Chapter 2.1 --- Semi-supervised learning --- p.6 / Chapter 2.2 --- Hierarchical Categorization --- p.10 / Chapter 3 --- Framework Overview --- p.13 / Chapter 4 --- Inherent Concept Detection --- p.18 / Chapter 4.1 --- Data Preprocessing --- p.18 / Chapter 4.2 --- Concept Detection Algorithm --- p.22 / Chapter 4.3 --- Kernel-based Distance Measure --- p.27 / Chapter 5 --- Training Example Discovery from Unlabeled Documents --- p.33 / Chapter 5.1 --- Training Document Discovery --- p.33 / Chapter 5.2 --- Automatically determining the number of extracted positive examples --- p.37 / Chapter 5.3 --- Classification Model --- p.39 / Chapter 6 --- Experimental Evaluation --- p.44 / Chapter 6.1 --- Corpus Description --- p.44 / Chapter 6.2 --- Evaluation Metric --- p.49 / Chapter 6.3 --- Result Analysis --- p.50 / Chapter 7 --- Conclusions and Future Work --- p.66 / Bibliography --- p.68 / Chapter A --- Detailed result on the inherent concept detection process for the TDT and RCV1 corpora --- p.73

Chan Yiu-Man. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2004. / Includes bibliographical references (leaves 123-127). / Abstracts in English and Chinese. / Abstract (in English) --- p.i / Abstract (in Chinese) --- p.iii / Acknowledgements --- p.iv / Contents --- p.v / List of Figures --- p.ix / List of Tables --- p.xi / Abbreviations --- p.xii / Chapter Chapter 1. --- Introduction --- p.1 / Chapter 1.1 --- Background --- p.1 / Chapter 1.1.1 --- Source plants of Radix Stemonae --- p.1 / Chapter 1.1.2 --- Medicinal usage of Radix Stemonae --- p.1 / Chapter 1.1.3 --- Stemonaceae --- p.2 / Chapter 1.1.4 --- Stemonaceae of China --- p.3 / Chapter 1.1.5 --- Circumscriptions of Stemonaceae --- p.3 / Chapter 1.1.6 --- Affinity of Stemonaceae --- p.7 / Chapter 1.2 --- Molecular Markers for Phylogenetic study --- p.10 / Chapter 1.2.1 --- Choosing appropriate DNA region(s) --- p.10 / Chapter 1.2.2 --- Chloroplast DNA markers --- p.10 / Chapter 1.2.3 --- Nuclear sequences --- p.11 / Chapter 1.3 --- Objectives --- p.13 / Chapter Chapter 2. --- Materials and Methods --- p.14 / Chapter 2.1 --- Sources of Samples and Their Treatment --- p.14 / Chapter 2.1.1 --- Fresh Materials --- p.14 / Chapter 2.1.2 --- DNA Samples --- p.15 / Chapter 2.1.3 --- Dried Medicinal Material from Commerical Market --- p.15 / Chapter 2.2 --- DNA Isolation from Plant Materials --- p.20 / Chapter 2.2.1 --- Reagents for DNA Isolation --- p.20 / Chapter 2.2.2 --- Procedures of DNA Isolation --- p.22 / Chapter 2.2.2.1 --- Treatments of Plant Materials --- p.22 / Chapter 2.2.2.2 --- CTAB (Cetyltrimethylammonium bromide) Method --- p.23 / Chapter 2.2.2.3 --- DNeasy ® Plant Mini Kit --- p.24 / Chapter 2.2.2.4 --- GenElute Plant Genomic DNA Miniprep --- p.24 / Chapter 2.2.2.5 --- Extraction method of Kang et. al (1998) --- p.25 / Chapter 2.2.2.6 --- Agarose Gel Electrophoresis of Genomic DNA --- p.26 / Chapter 2.3 --- Polymerase Chain Reaction (PCR) --- p.27 / Chapter 2.3.1 --- Reagents --- p.27 / Chapter 2.3.2 --- Procedures --- p.28 / Chapter 2.4 --- "Ligation, Transformation and Bacterial Culture for 5S rRNA Spacer Analysis" --- p.30 / Chapter 2.4.1 --- Reagents --- p.30 / Chapter 2.4.2 --- Procedures --- p.33 / Chapter 2.4.2.1 --- Ligation --- p.33 / Chapter 2.4.2.2 --- Transformation --- p.33 / Chapter 2.4.2.3 --- Blue-White Screening --- p.33 / Chapter 2.4.2.4 --- Plasmid Isolation --- p.34 / Chapter 2.4.2.5 --- Screening of plasmid DNA by PCR --- p.35 / Chapter 2.5 --- Cycle Sequencing and Electrophoresis --- p.36 / Chapter 2.5.1 --- Instruments and Reagents --- p.36 / Chapter 2.5.2 --- Procedures of Cycle Sequencing and Electrophoresis --- p.37 / Chapter 2.5.2.1 --- Cycle sequencing --- p.37 / Chapter 2.5.2.2 --- Ethanol Precipitation --- p.38 / Chapter 2.5.2.3 --- Electrophoresis --- p.38 / Chapter 2.6 --- Sequence Analysis --- p.40 / Chapter Chapter 3. --- Taxonomic Study of Chinese Stemona species --- p.41 / Chapter 3.1 --- History of the Genus Stemona --- p.41 / Chapter 3.2 --- Characteristics of the Genus Stemona --- p.42 / Chapter 3.3 --- Characteristics of Stemona sessilifolia (Miquel) Miquel (including Stemona shandongensis D. K. Zang) --- p.45 / Chapter 3.4 --- Characteristics of Stemona japonica (Blume) Miquel --- p.52 / Chapter 3.5 --- Characteristics of Stemona tuberosa Loureiro --- p.55 / Chapter 3.6 --- Characteristics of Stemona parviflora C. H. Wright --- p.61 / Chapter 3.7 --- Characteristics of Stemona mairei (H. Leveille) K. Krause --- p.65 / Chapter 3.8 --- Characteristics of Stemona kerrii Craib --- p.67 / Chapter Chapter 4. --- DNA Sequence Analysis for Authentication and Systematics --- p.69 / Chapter 4.1 --- DNA Extraction --- p.70 / Chapter 4.2 --- PCR --- p.73 / Chapter 4.3 --- DNA Authentication of Radix Stemonae --- p.77 / Chapter 4.3.1 --- TrnL intron sequences --- p.77 / Chapter 4.3.2 --- 5S rRNA spacer sequences --- p.86 / Chapter 4.3.3 --- Conclusion of DNA Authentication --- p.107 / Chapter 4.4 --- Molecular Systematics Analysis --- p.108 / Chapter 4.4.1 --- Circumscription of Stemonaceae and its affinity to other monocots based on trnL intron sequences --- p.109 / Chapter 4.4.2 --- Interspecific relationship of Stemona --- p.114 / Chapter Chapter 5. --- Discussion --- p.116 / Chapter 5.1 --- Molecular Authentication of Radix Stemonae --- p.116 / Chapter 5.2 --- Molecular Markers --- p.117 / Chapter 5.3 --- The Variation in Stemona tuberosa --- p.117 / Chapter 5.4 --- Comparsion of Stemona sessilifolia and S. shandongensis --- p.118 / Chapter 5.5 --- Circumscription of Stemonaceae --- p.119 / Chapter 5.6 --- Affinity of Stemonaceae --- p.120 / Chapter Chapter 6. --- Conclusion --- p.122 / References --- p.123

Master of Science / Department of Computer Science / William H. Hsu / Institutional attrition refers to the phenomenon of members of an organization leaving it over time - a costly challenge faced by many institutions. This work focuses on the problem of predicting attrition as an application of supervised machine learning for classification using summative historical variables. Raising the accuracy, precision, and recall of learned classifiers enables institutional administrators to take individualized preventive action based on the variables that are found to be relevant to the prediction that a particular member is at high risk of departure. This project focuses on using multivariate logistic regression on historical institutional data with wrapper-based feature selection to determine variables that are relevant to a specified classification task for prediction of attrition. In this work, I first describe a detailed approach to the development of a machine learning pipeline for a range of predictive analytics tasks such as anticipating employee or student attrition. These include: data preparation for supervised inductive learning tasks; training various discriminative models; and evaluating these models using performance metrics such as precision, accuracy, and specificity/sensitivity analysis. Next, I document a synthetic human resource dataset created by data scientists at IBM for simulating employee performance and attrition. I then apply supervised inductive learning algorithms such as logistic regression, support vector machines (SVM), random forests, and Naive Bayes to predict the attrition of individual employees based on a combination of personal and institution-wide factors. I compare the results of each algorithm to evaluate the predictive models for this classification task. Finally, I generate basic visualizations common to many analytics dashboards, comprising results such as heat maps of the confusion matrix and the comparative accuracy, precision, recall and F1 score for each algorithm. From an applications perspective, once deployed, this model can be used by human capital services units of an employer to find actionable ways (training, management, incentives, etc.) to reduce attrition and potentially boost longer-term retention.

No description available.

Several tools exist for root cause analysis (RCA). Despite this however, many practitioners are not obtaining the quality improvement desired. Those turning to literature for guidance would find most of the information resides in case studies with anecdotal outcomes. Since 5 Whys analysis has been one of the more pervasive tools in use, this study sought to add to the root cause analysis body of knowledge by investigating tool support factors. While studied in conjunction with 5 Whys, the support variables lend themselves to other root cause analysis tools as well. The purpose of the study was to utilize a 2 x 2 factorial design to determine the significance and effect on RCA effectiveness, of using a 5 Whys trained facilitator and action level classification. During the study, problem solving teams at service centers of a North American electric repair company conducted analysis with or without a trained facilitator. Additionally, corrective actions were or were not categorized by defined levels of ability to impact defect prevention. The dependent variable of effectiveness was determined by scoring from a weighted list of best practices for problem solving analysis. Analysis showed trained facilitators had significant effect on problem solving solutions, while classification had minimal

The idea of voting multiple decision rules was introduced in to statistics by Breiman. He used bootstrap samples to build different decision rules, and then aggregated them by majority voting (bagging). In regression, bagging gives improved predictors by reducing the variance (random variation), while keeping the bias (systematic error) the same. Breiman introduced the idea of bias and variance for classification to explain how bagging works. However, Friedman showed that for the two-class situation, bias and variance influence the classification error in a very different way than they do in the regression case. In the first part of the dissertation, we build a theoretical framework for ensemble classifiers. Ensemble classifiers are currently the best off-the-shelf classifiers available, and they are the subject of much current research in classification. Our main theoretical results arc two theorems about voting iid (independently identically distributed) decision rules. The bias consistency theorem guarantees that voting will not change the Bias set, and the convergence theorem gives an explicit rate of convergence. The two theorems explain exactly how ensemble classifiers work. We also introduce the concept of weak consistency as opposed to the usual strong consistency. A boosting theorem is derived for a distribution-specific situation with iid voting. In the second part of this dissertation, we discuss a special ensemble classifier called PERT. PERT is a voted random tree classifier for which each random tree classifies every training example correctly. PERT is shown to work surprisingly well. We discuss its consistency properties. We then compare its behavior to the NN (nearest neighbor) method and boosted c4.5. Both of the latter methods also classify every training example correctly. We call these types of classifiers “oversensitive” methods. We show that one reason PERT works is because of its “squeezing effect.” In the third part of this dissertation, we design simulation studies to investigate why boosting methods work. The outlier effect of PERT is discussed and compared to boosted and bagged tree methods. We obtain a new criterion (Bayes deviance) that measures the efficiency of a classification method. We design simulation studies to compare the efficiency of several common classification methods, including NN, PERT, and boosted tree method.