Last data update: 2014.03.03

 R: GlobalAncova with sequential and type III sum of squares...
GlobalAncova.decompR Documentation

GlobalAncova with sequential and type III sum of squares decomposition and adjustment for global covariates

Description

Computation of a F-test for the association between expression values and clinical entities. The test is carried out by comparison of corresponding linear models via the extra sum of squares principle. In models with various influencing factors extra sums of squares can be treated with sequential and type III decomposition. Adjustment for global covariates, e.g. gene expression values in normal tissue as compared to tumour tissue, can be applied. Given theoretical p-values may not be appropriate due to correlations and non-normality. The functions are hence seen more as a descriptive tool.

Usage

GlobalAncova.decomp(xx, formula, model.dat = NULL, method = c("sequential", "type3", "all"),  test.genes = NULL, genewise = FALSE, zz = NULL, zz.per.gene = FALSE)

Arguments

xx

Matrix of gene expression data, where columns correspond to samples and rows to genes. The data should be properly normalized beforehand (and log- or otherwise transformed). Missing values are not allowed. Gene and sample names can be included as the row and column names of xx.

formula

Model formula for the linear model.

model.dat

Data frame that contains all the variable information for each sample.

method

Whether sequential or type III decomposition or both should be calculated.

test.genes

Vector of gene names or a list where each element is a vector of gene names.

genewise

Shall the sequential decomposition be displayed for each single gene in a (small) gene set?

zz

Global covariate, i.e. matrix of same dimensions as xx.

zz.per.gene

If set to TRUE the adjustment for the global covariate is applied on a gene-wise basis.

Value

Depending on parameters test.genes, method and genewise ANOVA tables, or lists of ANOVA tables for each decomposition and/or gene set, or lists with components of ANOVA tables for each gene are returned.

Note

This work was supported by the NGFN project 01 GR 0459, BMBF, Germany.

Author(s)

Ramona Scheufele ramona.scheufele@charite.de
Reinhard Meister meister@tfh-berlin.de
Manuela Hummel hummel@ibe.med.uni-muenchen.de
Urlich Mansmann mansmann@ibe.med.uni-muenchen.de

See Also

Plot.sequential, pair.compare, GlobalAncova

Examples

data(vantVeer)
data(phenodata)
data(pathways)

# sequential or type III decomposition
GlobalAncova.decomp(xx = vantVeer, formula = ~ grade + metastases + ERstatus, model.dat = phenodata, method = "sequential", test.genes = pathways[1:3])
GlobalAncova.decomp(xx = vantVeer, formula = ~ grade + metastases + ERstatus, model.dat = phenodata, method = "type3", test.genes = pathways[1:3]) 

# adjustment for global covariate
data(colon.tumour)
data(colon.normal)
data(colon.pheno)
GlobalAncova.decomp(xx = colon.tumour, formula = ~ UICC.stage + sex + location, model.dat = colon.pheno, method = "all", zz = colon.normal)

Results


R version 3.3.1 (2016-06-21) -- "Bug in Your Hair"
Copyright (C) 2016 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

> library(GlobalAncova)
Loading required package: corpcor
Loading required package: globaltest
Loading required package: survival
> png(filename="/home/ddbj/snapshot/RGM3/R_BC/result/GlobalAncova/GlobalAncova.decomp.Rd_%03d_medium.png", width=480, height=480)
> ### Name: GlobalAncova.decomp
> ### Title: GlobalAncova with sequential and type III sum of squares
> ###   decomposition and adjustment for global covariates
> ### Aliases: GlobalAncova.decomp
> ### Keywords: models
> 
> ### ** Examples
> 
> data(vantVeer)
> data(phenodata)
> data(pathways)
> 
> # sequential or type III decomposition
> GlobalAncova.decomp(xx = vantVeer, formula = ~ grade + metastases + ERstatus, model.dat = phenodata, method = "sequential", test.genes = pathways[1:3])
$androgen_receptor_signaling
                  SSQ   df         MS         F             p
Intercept   42.463648   72 0.58977289 17.037286 7.721335e-190
grade       18.201029  144 0.12639604  3.651313  1.390422e-42
metastases   2.645268   72 0.03673984  1.061336  3.397694e-01
ERstatus    29.125115   72 0.40451549 11.685593 1.671092e-122
error      226.807952 6552 0.03461660        NA            NA

$apoptosis
                  SSQ    df         MS         F             p
Intercept   73.630998   187 0.39374865 11.912387 3.087910e-321
grade       28.756345   374 0.07688862  2.326172  5.282043e-40
metastases   7.186794   187 0.03843206  1.162715  6.428585e-02
ERstatus    45.666917   187 0.24420811  7.388220 1.799784e-172
error      562.475084 17017 0.03305372        NA            NA

$cell_cycle_control
                  SSQ   df         MS        F            p
Intercept   11.764576   31 0.37950246 9.212123 1.322430e-40
grade       15.890992   62 0.25630633 6.221634 1.478817e-44
metastases   2.028524   31 0.06543627 1.588414 2.073819e-02
ERstatus    10.863558   31 0.35043735 8.506590 9.911064e-37
error      116.213873 2821 0.04119598       NA           NA

Warning messages:
1: In anova.lm(lm(dummy.formula, model.dat)) :
  ANOVA F-tests on an essentially perfect fit are unreliable
2: In anova.lm(lm(dummy.formula, model.dat)) :
  ANOVA F-tests on an essentially perfect fit are unreliable
3: In anova.lm(lm(dummy.formula, model.dat)) :
  ANOVA F-tests on an essentially perfect fit are unreliable
> GlobalAncova.decomp(xx = vantVeer, formula = ~ grade + metastases + ERstatus, model.dat = phenodata, method = "type3", test.genes = pathways[1:3]) 
$androgen_receptor_signaling
                  SSQ   df         MS          F             p
Intercept   31.070643   72 0.43153671 12.4661789 2.060236e-132
grade        9.448638  144 0.06561554  1.8954937  9.546578e-10
metastases   1.568993   72 0.02179157  0.6295123  9.939806e-01
ERstatus    29.125115   72 0.40451549 11.6855933 1.671092e-122
error      226.807952 6552 0.03461660         NA            NA

$apoptosis
                  SSQ    df         MS        F             p
Intercept   33.073033   187 0.17686114 5.350719 3.067227e-107
grade       16.433991   374 0.04394115 1.329386  2.488033e-05
metastases   6.267846   187 0.03351789 1.014043  4.331773e-01
ERstatus    45.666917   187 0.24420811 7.388220 1.799784e-172
error      562.475084 17017 0.03305372       NA            NA

$cell_cycle_control
                  SSQ   df         MS        F            p
Intercept    4.173999   31 0.13464512 3.268404 3.399027e-09
grade        8.024304   62 0.12942426 3.141672 5.765491e-15
metastases   1.397575   31 0.04508307 1.094356 3.295908e-01
ERstatus    10.863558   31 0.35043735 8.506590 9.911064e-37
error      116.213873 2821 0.04119598       NA           NA

Warning messages:
1: In anova.lm(lm(dummy.formula, model.dat)) :
  ANOVA F-tests on an essentially perfect fit are unreliable
2: In anova.lm(lm(dummy.formula, model.dat)) :
  ANOVA F-tests on an essentially perfect fit are unreliable
3: In anova.lm(lm(dummy.formula, model.dat)) :
  ANOVA F-tests on an essentially perfect fit are unreliable
> 
> # adjustment for global covariate
> data(colon.tumour)
> data(colon.normal)
> data(colon.pheno)
> GlobalAncova.decomp(xx = colon.tumour, formula = ~ UICC.stage + sex + location, model.dat = colon.pheno, method = "all", zz = colon.normal)
$adjustment
               ssq df
adjustment 2017200  1

$sequential
                 SSQ    df        MS          F          p
Intercept  4779.2569  1747 2.7356937 10.6058621 0.00000000
UICC.stage  374.1633  1747 0.2141747  0.8303224 0.99999978
sex         431.1914  1747 0.2468182  0.9568760 0.88723374
location    484.2787  1747 0.2772059  1.0746844 0.02092398
error      3604.7348 13975 0.2579417         NA         NA

$typeIII
                 SSQ    df        MS         F          p
Intercept  2026.3657  1747 1.1599117 4.4967985 0.00000000
UICC.stage  461.6535  1747 0.2642550 1.0244760 0.24661362
sex         431.2039  1747 0.2468253 0.9569038 0.88708056
location    484.2787  1747 0.2772059 1.0746844 0.02092398
error      3604.7348 13975 0.2579417        NA         NA

Warning message:
In anova.lm(lm(dummy.formula, model.dat)) :
  ANOVA F-tests on an essentially perfect fit are unreliable
> 
> 
> 
> 
> 
> dev.off()
null device 
          1 
>


Created & Maintained by Osamu Ogasawara (osamu.ogasawara@gmail.com) and IMSBIO