I used kmplot.com to analyze the overall survival rate of the gene I'm interested in. When I analyzed gene A only, the P value was ns. When I analyzed gene B only, the P value was significant, with the higher expression group being the better outcome. But I used the ratio of 2 genes expression A/B, I got the following results (attached photo) which show the lower expression group is the better outcome. How can I interpret this graph?
See the attached paper for a discussion of this topic
David Booth
Hello Htoo Shwe Eain
Do this with Gehan-Breslow-Wilcoxon test. It is better to use Prism software instead of kmplot.com.
Look at equation 10.2 in the attached file.
Maybe, it is only an arithmetic trick
>When I analyzed gene A only, the P value was ns.
Let A=10 ( as a mean or median)
>When I analyzed gene B only, the P value was significant, with the higher expression group being the better outcome.
B=10 better outcome
B=1 worse outcome
>ratio of 2 genes expression A/B
10/10=1 better outcome
10/1=10 worse outcome
Low A/B is better than high A/B
Maybe I misunderstood something?
- Badges
- Science topic
- Similar topics
- Philosophy
- Esthetics
- Interpretation