Entering edit mode
Hi, I was running pathway analysis with a set of differentiated genes
on KEGG pathways by SPIA. I tried both SPIA package (version 2.14.0)
and graphite package (version 1.8.1). But I found the results from
these 2 packages are quite different. For example, the top 3 pathways
from the 2 packages are:
1. from SPIA package:
Name ID pSize NDE pNDE tA pPERT
pG pGFdr pGFWER Status
1 mTOR signaling pathway 04150 62 6 0.351094977 -17.835652 0.006
0.01508872 1 1 Inhibited
2 Cell cycle 04110 122 18 0.006213776 -12.637336 0.633
0.02571712 1 1 Inhibited
3 Fanconi anemia pathway 03460 47 8 0.027099800 7.615679 0.207
0.03468601 1 1 Activated
2. from graphite package:
Name pSize NDE pNDE
tA pPERT pG pGFdr pGFWER Status
1 Measles 80 6 0.20253998
-26.44386 0.018 0.02411352 0.9830722 1 Inhibited
2 Chemokine signaling pathway 148 8 0.44861958
43.06511 0.010 0.02874194 0.9830722 1 Activated
3 Cytokine-cytokine receptor interaction 199 15 0.06820431
14.58672 0.093 0.03844114 0.9830722 1 Activated
The mTOR pathway that was the top pathway from SPIA package analysis
had very different results when analyzed by graphite package:
Name pSize NDE pNDE tA pPERT pG
pGFdr pGFWER Status
109 mTOR signaling pathway 53 1 0.932103 0 1 0.997641
0.9999985 1 Inhibited
So is this due to different versions of KEGG databases used in these 2
packages? It appears that mTOR pathway had fewer nodes (53) in
graphite package than in SPIA package (62), does that indicate KEGG
database is older in graphite package than in SPIA package? The
DESCRIPTION file of graphite package indicated the date of version
1.8.1 was 2013-11-03, so appears to be pretty new....
Thanks, any suggestions are appreciated.
John
[[alternative HTML version deleted]]
