Introduction Methodology The IsoGeneGUI Package Availability

IsoGeneGUI: a graphical user interface foranalyzing dose-response studies in microarray

experiments

Setia Pramana

Interuniversity Institute for Biostatistics and StatisticalBioinformatics, Universiteit Hasselt, Diepenbeek, Belgium

Introduction Methodology The IsoGeneGUI Package Availability

Outline

IntroductionDose-response StudiesIsoGene PackageIsoGeneGUI Package

MethodologyTest Statistics for Trend TestAsymptotic E2

PermutationSignificance Analysis of Microarray (SAM)

The IsoGeneGUI PackageMenuData ExplorationPermutationLikelihood Ratio Test E2 AnalysisSAMPlots

Availability

Introduction Methodology The IsoGeneGUI Package Availability

Dose-response studies

The fundamental objective of drug development.

Aims: find dose or range of dose that efficacious and safe.

Introduction Methodology The IsoGeneGUI Package Availability

Dose-response studies

To carry out the analysis of dose-response microarrayexperiments, Lin et al. (2007) discused the following teststatistics which take into account the order restriction of meanswith respect to the increasing doses:

Williams (Williams, 1971, 1972)

Marcus (Marcus, 1976)

the likelihood ratio test (E2) (Bartholomew 1961 , Barlow etal. 1972, and Robertson et al. 1988)

M statistic (Hu et al. 2005)

Modified M statistic (Lin et al. 2007)

Introduction Methodology The IsoGeneGUI Package Availability

IsoGene Package

An R package that carries out analysis of dose-responsemicroarray experiments discussed by Lin et.al 2007.

The inference is based on resampling methods, bothpermutations (Ge et al., 2003) and the SignificanceAnalysis of Microarrays (SAM, Tusher et al., 2001).

To control the False Discovery Rate (FDR) the BenjaminiHochberg (BH) procedure is implemented.

Available at: http://cran.ii.uib.no/web/packages/IsoGene/

Introduction Methodology The IsoGeneGUI Package Availability

IsoGeneGUI Package

The interface of the IsoGene package.

A graphical user interface based on R-tcl/tk package(Dalgaard, 2001).

A menu based package and data analysis can beperformed simply by selecting options from the menus ofthe package

Produces some default graphical displays as well asuser-defined graphical output.

Introduction Methodology The IsoGeneGUI Package Availability

Test Statistics for Trend Test

Test statistic Formula

Likelihood Ratio E201 =

∑

ij (yij−µ)2−

∑

ij (yij−µ⋆

i )2

∑

ij (yij−µ)2

Test (LRT)

Williams t = (µ⋆

K − y0)/√

2 ×

∑Ki=0

∑nij=1(yij − µi )2/(ni (n − K ))

Marcus t = (µ⋆

K − µ⋆

0 )/√

2 ×

∑Ki=0

∑nij=1(yij − µi )2/(ni (n − K ))

M M = (µ⋆

K − µ⋆

0 )/√

∑Ki=0

∑nij=1(yij − µ⋆

i )2/(n − K )

Modified M (M’) M′ = (µ⋆

K − µ⋆

0 )√

∑Ki=0

∑nij=1(yij − µ⋆

i )2/(n − I)

Introduction Methodology The IsoGeneGUI Package Availability

Analysis

Three analyses can be performed in the package:

E201 using its null distribution.

The five test statistics using permutations.

The Significance Analysis of Microarrays.

Introduction Methodology The IsoGeneGUI Package Availability

Likelihood ratio test (E2)

The null distribution of E2 test statistic is given by

P[E201(v) ≥ c] =

k∑

l=1

P(l , k ; w)P[B(l−1)/2,(v+k−l)/2 ≥ c], (1)

where l = 1, , 2, ..., k , v = N − k , Ba,b is a beta variable withparameters a and b, w is the weight vector through the levelprobabilities P(l , k ; w) as defined in Robertson et al. 1988.

Introduction Methodology The IsoGeneGUI Package Availability

Permutation

By permuting the labels of arrays randomly, the permutationtest statistics of m genes are re-calculated and obtain thepermutation matrix T :

T =

t11 t11 . . . t1B

t21 c22 . . . t2B...

...tm1 tm2 . . . tmB

(2)

where B is the number of permutations and each element tib ofmatrix T is the test statistic for the i-th gene in the b-thpermutation.The raw p-values:

Pi =#(b : |tib| ≥ |ti |)

B − 1(3)

where ti is the observed test statistic for gene i (Ge et al. 2003).

Introduction Methodology The IsoGeneGUI Package Availability

Significance Analysis of Dose-response MicroarrayData (SAM, Tusher et al., 2001)

The generic algorithm of SAM discussed by Chu et al.(2001) is implemented in this package.The SAM regularized test statistic:

For the t-type test statistics (i.e., Williams, Marcus, the M,and the M ′), a fudge factor is added in the standard error ofthe mean difference. For example:

M ′SAM =µ

⋆

K − µ⋆

0

s′ + s0, (4)

where s0 is the fudge factor.For the F -type test statistic, such as E2

01, is defined by

E2SAM01 =

√

σ2H0

− σ2H1

√

σ2H0

+ s0

. (5)

Introduction Methodology The IsoGeneGUI Package Availability

The IsoGeneGUI

Introduction Methodology The IsoGeneGUI Package Availability

The IsoGeneGUI: Menu

FileThis menu is used for loading and showing the data.Sub-menus:

Open DataR workspace (*.RData files)Excel or text file (*.xls or *.txt files)

Show DataExit

AnalysisThis is the main menu where the analyses are performed.Sub-menus:

Set seedLikelihood Ratio Test E2 AnalysisPermutation AnalysisSignificant Analysis of Microarrays (SAM)

SAM PermutationSAM Analysis

Introduction Methodology The IsoGeneGUI Package Availability

The IsoGeneGUI: Menu

PlotsBesides plots that can be produced in each the analysisdialog box, the package also provides extra plots:

IsoPlotPermutation PlotSAM Plot

Plot of FDR vs. DeltaPlot of number of significant genes vs. DeltaPlot of number of False Positive vs. Delta

User defined scatter plot

Help

IsoGene HelpIsoGeneGUI Manual and ExamplesAbout

Introduction Methodology The IsoGeneGUI Package Availability

Data Exploration

The IsoPlot dialog box

Introduction Methodology The IsoGeneGUI Package Availability

Data Exploration

The Isotonic regression plots The summary statistic

Introduction Methodology The IsoGeneGUI Package Availability

Likelihood Ratio Test E2 Analysis

The dialog box

Introduction Methodology The IsoGeneGUI Package Availability

Likelihood Ratio Test E2 Analysis

Output: List significant genes

Can be saved as R object or/and an excel file.

Introduction Methodology The IsoGeneGUI Package Availability

Likelihood Ratio Test E2 Analysis

Output: Fold change vs. E2 plot

Can be copied into clipboard and be saved as different imagetypes.

Introduction Methodology The IsoGeneGUI Package Availability

Permutation

Introduction Methodology The IsoGeneGUI Package Availability

Permutation

The output:

Result for all genes.

list of significant genes using the selected statistic(s).

Graphical displays.

Introduction Methodology The IsoGeneGUI Package Availability

Significance Analysis of Microarrays (SAM)

The SAM permutation dialog box

Introduction Methodology The IsoGeneGUI Package Availability

Significance Analysis of Microarrays (SAM)

Introduction Methodology The IsoGeneGUI Package Availability

Significance Analysis of Microarrays (SAM)

Output: Delta table

Introduction Methodology The IsoGeneGUI Package Availability

Plots

Permutation p-values plot

Introduction Methodology The IsoGeneGUI Package Availability

Plots

User Defined Scatter Plot

Introduction Methodology The IsoGeneGUI Package Availability

Help

Html help with screenshots

Introduction Methodology The IsoGeneGUI Package Availability

Infobox

Provides information about the data (availability and summary)and the result summary of the last performed analyses.

Introduction Methodology The IsoGeneGUI Package Availability

Availability

R-forge site:https://r-forge.r-project.org/projects/isogenegui/

To install: install.packages(”IsoGeneGUI”,repos=”http://R-Forge.R-project.org”)

The full user manual and example data will be soonavailable at: http://www.censtat.uhasselt.be/software/

Plan to be submited to Bioconductor

Introduction Methodology The IsoGeneGUI Package Availability

Selected References

Barlow, R.E., Bartholomew, D.J., Bremner, M.J. and Brunk, H.D.(1972) Statistical Inference Under Order Restriction, New York:Wiley.

Benjamini, Y. and Hochberg, Y. (1995) Controlling the falsediscovery rate: a practical and powerful approach to multipletesting, J. R. Statist. Soc. B, 57, 289-300.

Lin, Dan, Shkedy, Ziv, Yekutieli, Dani, Burzykowski, Tomasz,Gohlmann, Hinrich, De Bondt, An, Perera, Tim, Geerts, Tamaraand Bijnens, Luc.(2007) Testing for Trends in Dose-ResponseMicroarray Experiments: A Comparison of Several TestingProcedures, Multiplicity and Resampling-Based Inference,Statistical Applications in Genetics and Molecular Biology: Vol. 6: Iss. 1, Article 26.

Robertson, T., Wright, F.T. and Dykstra, R.L. (1988), OrderRestricted Statistical Inference, Wiley.

Introduction Methodology The IsoGeneGUI Package Availability

Thank you for your attention...

Demo??