Scicraft:Affymetrics workflow
From Scicraft wiki
Contents |
Affymetrics
Here we describe the usage of the functions in the Affymetrics [wiki:workflow workflow] included in SciCraft.
Load Data stage
Import Affy
To start working with your Affymetrics CEL files, you use this function to import them into SciCraft. You load multiple files and the returned result is a 2d-array of genes to files. To select multiple files, either hold down the CTRL-Key and click additional files, or hold down the SHIFT-key to select a range.
Import Phenotypes
This function allow you to import a description of your phenotypes created in Gnumeric/Excel or a similar spreadsheet application. The file must have been saved as tab-delimited text, such as the following example:
{{{ CEL Child age sex infected 02-29-199 199 8 F I 02-29-3047 3047 8 M I 02-29-3433 3433 8 M I 02-29-1291 1291 6 F N 02-29-3182 3182 8 F N 02-29-3422 3422 7.5 M N }}}
[attachment:covaraiates.txt Downloadable example.]
Significance Analysis
LIMMA
The LIMMA (Linear Models for Microarray Data) function uses the LIMMA library works on two datasets, the affymetrics dataset (as imported by Import Affy), and the phenotype description (as imported by Import Phenotypes). You can specify a list of comparisons to make between the categories of the phenotype description.
This analysis results in a Vulcano Plot and datasets for coefficients, average intensity and adjusted p-values.
For more information about LIMMA, see: http://bioinf.wehi.edu.au/limma/
Explorative Analysis
PCA
Performs a PCA (principal components analysis) on a dataset using the R function prcomp.
Returns T and P matrices together with loadings and scores plots.
Plot Data
Scatterplot
Create a scatter plot for one dataset or plot two datasets against eachother. To plot two datasets they need to have a shared dimension (you can't plot apples against oranges).
Save Data
Export Dataset(s)
This function allows you to export one or more datasets as a tab-delimited textfile, including row and column headers.
Exporting multiple datasets at the same time will join the datasets along a shared dimension.
