A Novel Application of Regression Analysis (MM-Estimator) with Simultaneous Prediction Confidence Intervals are proposed to detect up- or down-regulated genes, which are outliers in scatter plots based on log-transformed red (Cy5 fluorescent dye) versus green (Cy3 fluorescent Dye) intensities. Advantages of the application: 1) Robust and Resistant MM-Estimator is a Reliable Method to Build Linear Regression In the presence of Outliers, 2) Exploratory Data Analysis Tools (Boxplots, Averaged Shifted Histograms, Quantile-Quantile Normal Plots and Scatter Plots) are Unsed to Test Visually underlying assumptions of linearity and Contaminated Normality in Microarray data), 3) Simultaneous prediction confidence intervals (SPCIs) Guarantee a desired confidence level across the whole range of the data points used for the scatter plots. Results of the outlier detection procedure is a set of significantly differentially expressed genes extracted from the employed microarray data set. A scatter plot smoother (super smoother or locally weighted regression) is used to quantify heteroscendasticity is residual variance (Commonly takes place in lower and higher intensity areas). The set of differentially expressed genes is quantified using interval estimates for P-values as a probabilistic measure of being outlier by chance. Monte Carlo simultations are used to adjust super smoother-based SPCIs.her.
To order this software or receive further information, please fill out the following request: Request Software
Loguinov, Alexandre V. (2002, July 19). MM-Estimator and Adjusted Super Smoother based Simultaneous Prediction Confedenc (Version 00) [Computer software].
@misc{osti_1230999,
title = {MM-Estimator and Adjusted Super Smoother based Simultaneous Prediction Confedenc, Version 00},
author = {Loguinov, Alexandre V.},
abstractNote = {A Novel Application of Regression Analysis (MM-Estimator) with Simultaneous Prediction Confidence Intervals are proposed to detect up- or down-regulated genes, which are outliers in scatter plots based on log-transformed red (Cy5 fluorescent dye) versus green (Cy3 fluorescent Dye) intensities. Advantages of the application: 1) Robust and Resistant MM-Estimator is a Reliable Method to Build Linear Regression In the presence of Outliers, 2) Exploratory Data Analysis Tools (Boxplots, Averaged Shifted Histograms, Quantile-Quantile Normal Plots and Scatter Plots) are Unsed to Test Visually underlying assumptions of linearity and Contaminated Normality in Microarray data), 3) Simultaneous prediction confidence intervals (SPCIs) Guarantee a desired confidence level across the whole range of the data points used for the scatter plots. Results of the outlier detection procedure is a set of significantly differentially expressed genes extracted from the employed microarray data set. A scatter plot smoother (super smoother or locally weighted regression) is used to quantify heteroscendasticity is residual variance (Commonly takes place in lower and higher intensity areas). The set of differentially expressed genes is quantified using interval estimates for P-values as a probabilistic measure of being outlier by chance. Monte Carlo simultations are used to adjust super smoother-based SPCIs.her.},
doi = {},
url = {https://www.osti.gov/biblio/1230999},
year = {Fri Jul 19 00:00:00 EDT 2002},
month = {Fri Jul 19 00:00:00 EDT 2002},
note =
}
