Statistical Tests Exercise (Wed Nov 23 2006)
   ============================================
   Follow-up course of Microarray Data Analysis
   Nov 20-24 2006, PICB Shanghai
   Christine Steinhoff, Dennis Kostka

1. Differential Expression.
   Load the file /picb/course/50-Statistical-Test/ALL.data.txt
   into TIGR MultiExperiment Viewer (MeV). It contains data from two independant
   groups of patients. Which t-Test is applicable? Perform it, with a
   significance level (critical p-value) alpha = 0.05.
   How many Genes are shown as significantly expressed between the
   two groups?

2. Multiple Testing.
   Correct the p-values for multiple testing. Compare the standard
   Bonferroni correction with the adjusted version. Also use the 
   permutation procedure for p-value calculation. Try different 
   numbers of permutations. 

3. False Discovery Rate.
   How many "significant" genes do you get when accepting 
   (with a confidence of 95%) ten false positives amongst them?
   Confidence means, that the probability of more than ten
   false positives is smaller than 5%.

4. Heat map.
   Construct a hierarchical clustering tree for one of the groups of 
   significant genes. Cluster the genes only, not the samples.

5. Perform these analysis steps also for the data sets: 
   /picb/course/50-Statistical-Test/data.lung.txt  (lung cancer)
   /picb/course/50-Statistical-Test/foodat.txt  (random noise)
   
6. Compare results for the random and the lung data.