Murine atherosclerosis is induced by a high fat diet in susceptible mice with either APoE-/- or LDLr-/- in c57 background. Sections are taken from the aortic sinus to quantify area of plaque in total sinus area. Form each mice a minimum of 6 sections are quantified. If we have 6 animals per group, how do we carry out statistical analysis? Average of individual mice and take N=6 or take individual section N= 6X6=36
Can anyone tell me which is the correct method? Does the same rule apply to immunostaining too?
It all depend how you have collected the sections? there is good paper by Geoff Werstuck and I am Co author on that and that will tell you about the sectioning and collection. That way you will have very close matching sections from each mice and then you can compare by average per group. read the paper, if still have let me know
I would use nonparametric Mann-Whitney test, where APoE-/- and LDLr-/- will be a factor. (so 72samples with 2 different names of factor) then you do not have to average anything, plus (compare to parametric tests such as ANOVA) your results will not depend on assumptuion of normal distribution and equal variances. Disadvantage of such test (compare to ANOVA) is that you will not know if there are differences between individual mice in the same group, but it is not your goal if I understood you correctly.
Here is link for the paper
"Quantitative analysis and characterization of atherosclerotic lesions in the murine aortic sinus.
Venegas-Pino DE, Banko N, Khan MI, Shi Y, Werstuck GH.
J Vis Exp. 2013 Dec 7;(82):50933. doi: 10.3791/50933."
here is video too
Hi
Thanks for the details.
Unfortunately we do not have a subscription for this journal. Can you please share the full text.
Thanks
Lakshmi,
Your statistical problem has nothing to do with what you measure (i.e., atherosclerosis) but with your methodology:
- If I understand your methodology you want to compare 2 parallel groups;
- as you have 6 measurements by mouse I advice you using the mean of the 6 measurements in each individual for your variable;
- I agree with Stanislav that the best test is the Mann-Whitney test (i.e., when comparing 2 parallel groups) as your sample is quite small (n=6);
- however your sample is far too small as you can only compare 2 parallel groups of n=3 in each group !!!
I performed lesion measurement in the aortic root all the time. Normally for each animal I have 10 sections and in each section I sum all lesions area and then I take the average of the 10 section lesion areas. At the end if you have 6 animals the you should have 6 averages of the summation which you can use for statistical analysis. Hope that helps.
- Badges
- Science topic
- Similar topics
- Medicine
- Disease
- Disease Models
- Animal Models
- Rodent Models
- Mouse Models