Single-, two-, multi-class SAM

These SAM versions are extensions of

• Single class SAM => Single class T-test
• Two class SAM => Two class T-test
• Multi-class SAM => ANOVA

Single Class SAM

In this version of SAM genes are searched which show clear regulation throughout all selected hybridisdations.

Basic test statistic is a single class T-test.

Parameters:

• One group of hybridisations has to be defined
• Number of permutations
• How to find s₀ (fudge factor)
• Delta

Results:

• Delta table
• Values for selected Delta
• List of up-regulated significant genes
• List of down-upregulated significant genes
• List of regulated significant genes
• List of non-regulated significant genes

Gene lists may be visualised and saved as

• genelist with basic statistical values
• sub-expression matrices
• expression profiles
• centroid graphs
• heatmaps

Two class SAM

In this version of SAM genes are searched which can distinguish between two groups of hybridisations (e.g. cancer <=> normal).

Basic test statistic is a two class T-test.

Parameters:

• Two groups of hybridisations have to be defined
• Number of permutations
• Permutation balancing
• How to find s₀ (fudge factor)
• Delta

Results:

• Delta table
• Values for selected Delta
• List of up-regulated significant genes
• List of down-upregulated significant genes
• List of regulated significant genes
• List of non-regulated significant genes

Gene lists may be visualised and saved as

• genelist with basic statistical values
• sub-expression matrices
• expression profiles
• centroid graphs
• heatmaps

Multi class SAM

In this version of SAM genes are searched which can distinguish between three or more sub-groups of hybridisations
(e.g. kidney <=> liver <=> pancreas <=> lung <=> brain <=> ...).
The present version can handle a maximum of 12 independant groups.

Basic test statistic is ANOVA

Parameters:

• Three or more groups of hybridisations have to be defined
• Number of permutations
• Permutation balancing
• How to find s₀ (fudge factor)
• Delta

Results:

• Delta table
• Values for selected Delta
• List of up-regulated significant genes
• List of down-upregulated significant genes
• List of regulated significant genes
• List of non-regulated significant genes

Gene lists may be visualised and saved as

• genelist with basic statistical values
• sub-expression matrices
• expression profiles
• centroid graphs
• heatmaps

In  the SAM graph you find only the upper right quadrant (both x-  / and y-axis have only positive values).
Also, you only find selections for positive significant and non significant genes.

last edited 26.04.2006