log4.txt

===============================================================================

2 September 2015

-------------------------------------------------------------------------------

Turns out that I needed to log the data before quality control/removing the bad samples.
    - Re-do the analysis to see if I get different results

Some data points had 0 counts which caused some problems with logging
    - added 1 to all the count data to avoid this

After logging the data, there were no samples that had a correlation of 0.6
    - No samples were discarded

SVD on Creighton et al. data was done and used to transform the endometrial data
    - The results didn't change
        - Metagene from Creighton et al data doesn't correlate with the endometrial data...

Do DEG analysis on endometrial data and compare this with obesity genes identified in Creighton et al. data.