I am using GraphPad prism for data visualisation and analysis. I am trying to assess if there is a statistically significant difference between the percentage of 2 cohorts (patients and cell lines) that have the same gene mutated. I identified the top 10 genes mutate din the patient group and noted down the percentage of the group that had each mutated gene. I used this, almost as a control group, and identified the percentage of cell lines that have those same genes mutated. Originally, GraphPad suggested 2 way ANOVA, but on further research I believe this to be the incorrect test. It seems that Fisher's exact test would be most appropriate, but I'm having issues running this on GraphPad as the test is taking forever to run. I also want to do the same analysis for 3 cohorts, would the test be different. I would be grateful for all help and signposting as this is for my dissertation, due in a few days. I have attached pictures of the graphs for context.
Even if some calculated value is obtained using statistical analysis software, it is meaningless. There is nothing to be said from Asian n=2 sample. Even if you could calculate something, it would be worthless. You should remove the Asian sample from the statistical analysis, and only do the statistical analysis for the Black/White comparison, and display Asian separately on the graph.
- Badges
- Science topic
- Similar topics
- Filing