Investigator: Chad Johnson, Beachaump Lab

Description: 16 treatment pairs, pre- and 7 days post-treatment with Celebrex. Comparison of gene expression levels between these treatment pairs.

Tasks

Normalization (Zhouwen Liu).

Used RMA (in Biocondutor package) to normalize these data. Completed 9/20/05.

Permutation paired t-test (Zhouwen Liu).

Used AffyPaired software. Completed 9/21/05. There were 202 probes with differential experession p<0.01, and 15 probes with p<0.001. Decided to use 202 probe set for annotation and further examination of pathways.

Annotation (Aixiang Jiang).

Used Bioconductor GOTERMS package. Completed 11/17/05.

After meeting with Chad, it was decided that we would try to use KEGG terms instead. Tried the globaltest function - currently not functional for use with paired data (point to work on); no statistically significant pathways using this function. Most likely not powerful enough with small sample size (and we lose any power than might have been gained by using paired experimental design).

Report of 34 genes (those with unique LocusLink id's), with KEGG term description and direction of change was given to Chad Johnson on 1/18/06.

Additional requests

email from 1/28/06

To that end, could you please tell me which direction the following two genes went (with regard to celecoxib treatment status) in our set? I don't have their Affy probe IDs, but these locus links were derived from the gene set you gave me, so I know they're in our list. The Locus Link IDs are:

• 200895_s_at 2288 FKBP4 FK506 binding protein 4, 59kDa
• 210187_at 2280 FKBP1A FK506 binding protein 1A, 12kDa
• 214119_s_at 2280 FKBP1A FK506 binding protein 1A, 12kDa
• 219249_s_at 60681 FKBP10 FK506 binding protein 10, 65 kDa
• 235311_at 55033 FKBP14 FK506 binding protein 14, 22 kDa
• 228006_at 5728 PTEN phosphatase and tensin homolog (mutated in multiple advanced cancers 1)

Bonnie finished this search and sent email (2/2/06) - all probes were higher in PRE with the exception of 228006_at which was higher in POST.