Test for significance? - (May/08/2006 )

I'm testing the ability of mutant proteins to complement protein-null viruses. I've got a long list of average PFUs to compare to the wild type protein. How do I compare them? I don't know how to compare the PFUs for these mutants.

For example, my wild type virus gave a PFU of 3.18 X 10E7
I have a mutant that gave a PFU of 2.43 X 10E7
and another that gave a PFU of 1.8 X 10E4

Clearly, the first mutant is behaving similar to the wild type protein, and the second mutant is significantly different, but I don't know how to calculate those significance values using such huge numbers. Any formula I use gives me huge p values, due to having such large values for PFU. These numbers are averages of three separate experiments.

Does anyone have any suggestions. Please, help!

Thanks a lot.

You can go lots of places on the web where you can calculate the p-values for given values. Some examples:

http://www.stat.sc.edu/~ogden/javahtml/pvalcalc.html
http://faculty.vassar.edu/lowry/tabs.html

an Excel Addin:
http://www.wabash.edu/econometrics/Econome...eCalculator.htm

Or use a statistics software like SPSS or SigmaStat.

In every case you will need the raw data of your three experiments, where three times might be a little low on the n side of the tests, which may likely give you high (>0.1) p-values, so you maybe should consider additional reproductions of the experiment (like n=5).

Mike

Try students t test and enter all the values you have for the different PFU. Dont average it out yourself, let the program do it for u.

http://www.physics.csbsju.edu/stats/t-test.html

http://www.graphpad.com/quickcalcs/ttest1.cfm

these sites should help you.