Hello,
I'm a medicine student and i'm investigating the effects of a single nucleotide polymorphism on the expression of a phenotype (cuantitative).
I have already performed an ANOVA test to see if there where any significant differences between the groups WT/WT, WT/VT and VT/VT. Out of nearly 200 genes I found 3 genes with statistical significance.
I would continue to perform tests on these genes in order to find out more about the relationship between these genes and the phenotype.
What do you think should be the next step in order to find the impact of having 0, 1 or 2 copies of the allele VT on the phenotype?
What test would you perform and with which data would you present the discoveries to a scientific journal?
Thank you very much in advance.
Hello Fernando Huet , you have two choices to compare your quantitative variable between the groups for each gene:
- Comparison of the means : Student's t-test if the quantitative variable follows a normal low, or a non parametric test (e.g. Mann-Witney test) if not.
- Test of correlation : Pearson correlation test if the quantitative variable follows a normal law, or a Spearman correlation test if not.
In the scientific journals you can present the P-values resulting of these test, and for better quality add the size effect.
If you have a question you can ask me. Good luck.
From your explain you have two allele and three genotypes and phenotypes of investigated individuals. You want to detect the relationship between genotypes and phenotypes.
You are right when used ANOVA to detect, you also used ANOVA if there are more than three genetype. I recommend you to use any source of variation may affect the phenotype in your statistical model this method may be decreased error and will increase the degree of significance.
Also you can determine the Average Gene Effect and Gene Substitution from genotypes means. When you calculate the Average Effect of each allele (as addative value), then you can estimated dominance deviation.
Ibrahim Ishag thank you for your answer. What test in particular would you use to determine de Average Gene Effect? Would the Pearsons Correlation Test be enough?
Please find to the attached files. The first one give theory of average gene affect while the second give you implication of it in SNP may similar to your work.
- Badges
- Science method