Comparing ASCA and clustering identified proteins
I modified and reran the clustering code using the technical replicate averages. The code and other plots can be found here. I used the proteins ASCA identified as being affected by temperature.
This is the code I used to compare the analyses. A total of 25 proteins were common between clustering and ASCA analysis. ASCA identified 113 different proteins while clustering identified 8 other proteins. It would be interesting to see how manipulating the dendrogram height cutoff value during clustering affects the latter number (I used a cutoff of 250).
Here is a heatmap of the similar proteins (note the first column is Day 0 followed by 23C days 3-13 and finally 29C days 3-13):
Venn Diagram of 23C and 29C
I’m working on making a better venn diagram in R with just the silo 3 and 9 proteins. Here is a quick and dirty venn digram, but I’m trying to usethe VennDiagram package to improve it.
Note that all 0 values were changed to 0.100 in the technical replicate averages. I removed all values less than 0.600 as these would be undetected in the silo because of the change (there are 6 days of data since were are excluding day 15 because of the temperature malfunction).
This is obviously still a work in progress to create a better venn diagram.