Evaluating test statistics to select interesting genes in microarray experiments.

Charles Kooperberg; Simonetta Sipione; Michael LeBlanc; Andrew D Strand; Elena Cattaneo; James M Olson

doi:10.1093/hmg/11.19.2223

Home/Archive/PubMed

PubMed · 12217950

Evaluating test statistics to select interesting genes in microarray experiments.

Charles Kooperberg·Simonetta Sipione·Michael LeBlanc·Andrew D Strand·Elena Cattaneo·James M Olson

View original source ↗Export metadata ↓

Abstract

A randomization procedure to evaluate the significance level and the false-discovery rate in complex microarray experiments is proposed. A related graph can be used to compare different test statistics that can be used to analyze the same experiment. This graph is closely related to receiver operator characteristic (ROC) curves. The proposed method is applied to a subset of the data from a cell-line experiment related to Huntington's disease. A small simulation study compares the effectiveness of the proposed procedure with the significance analysis of microarrays (SAM) procedure.

Explore related subjects

Computer Simulation Data Interpretation, Statistical Huntington Disease Models, Genetic Oligonucleotide Array Sequence Analysis ROC Curve Random Allocation Regression Analysis Journal Article Research Support, Non-U.S. Gov't Research Support, U.S. Gov't, P.H.S.

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Charles Kooperberg, Simonetta Sipione, Michael LeBlanc, Andrew D Strand, Elena Cattaneo, James M Olson. 2002-09-15. Evaluating test statistics to select interesting genes in microarray experiments.. https://doi.org/10.1093/hmg%2F11.19.2223

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2025

Generating correlated data for omics simulation.

Simulation of realistic omics data is a key input for benchmarking studies that help users obtain optimal computational pipelines. Omics data involves large numbers of measured features on each sample and these measures are generally correlated with each other. However, simulation too often ignores these correlations, perhaps due to computational and statistical hurdles of doing so. To alleviate this, we describe three approaches for generating omics-scale data with correlated measures which mimic real datasets. These approaches are all based on a Gaussian copula approach with a covariance matrix that decomposes into a diagonal part and a low-rank part. This decomposition allows for extremely efficient simulation, overcoming a hurdle for adoption of past methods. We use these approaches to demonstrate the importance of including correlation in two benchmarking applications. First, we show that variance of results from the popular DESeq2 method increases when dependence is included. Second, we demonstrate that CYCLOPS, a method for inferring circadian time of collection from transcriptomics, improves in performance when given gene-gene dependencies in some circumstances. We provide an R package, dependentsimr, that has efficient implementations of these methods and can generate dependent data with arbitrary marginal distributions, including discrete (binary, ordered categorical, Poisson, negative binomial), continuous (normal), or with an empirical distribution.

Computer Simulation↗

PubMed2017

Addressing current challenges in cancer immunotherapy with mathematical and computational modelling.

The goal of cancer immunotherapy is to boost a patient's immune response to a tumour. Yet, the design of an effective immunotherapy is complicated by various factors, including a potentially immunosuppressive tumour microenvironment, immune-modulating effects of conventional treatments and therapy-related toxicities. These complexities can be incorporated into mathematical and computational models of cancer immunotherapy that can then be used to aid in rational therapy design. In this review, we survey modelling approaches under the umbrella of the major challenges facing immunotherapy development, which encompass tumour classification, optimal treatment scheduling and combination therapy design. Although overlapping, each challenge has presented unique opportunities for modellers to make contributions using analytical and numerical analysis of model outcomes, as well as optimization algorithms. We discuss several examples of models that have grown in complexity as more biological information has become available, showcasing how model development is a dynamic process interlinked with the rapid advances in tumour-immune biology. We conclude the review with recommendations for modellers both with respect to methodology and biological direction that might help keep modellers at the forefront of cancer immunotherapy development.

Computer Simulation↗

PubMed2003

Scale-down of continuous filtration for rapid bioprocess design: Recovery and dewatering of protein precipitate suspensions.

The early specification of bioprocesses often has to be achieved with small (tens of millilitres) quantities of process material. If extensive process discovery is to be avoided at pilot or industrial scale, it is necessary that scale-down methods be created that not only examine the conditions of process stages but also allows production of realistic output streams (i.e., streams truly representative of the large scale). These output streams can then be used in the development of subsequent purification operations. The traditional approach to predicting filtration operations is via a bench-scale pressure filter using constant pressure tests to examine the effect of pressure on the filtrate flux rate and filter cake dewatering. Interpretation of the results into cake resistance at unit applied pressure (alpha) and compressibility (n) is used to predict the pressure profile required to maintain the filtrate flux rate at a constant predetermined value. This article reports on the operation of a continuous mode laboratory filter in such a way as to prepare filter cakes and filtrate similar to what may be achieved at the industrial scale. Analysis of the filtration rate profile indicated the filter cake to have changing properties (compressibility) with time. Using the insight gained from the new scale-down methodology gave predictions of the flux profile in a pilot-scale candle filter superior to those obtained from the traditional batch filter used for laboratory development.

Computer Simulation↗