p*****V 发帖数: 43 | 1 I am doing genetic association study. In order to adjust for ethnicity
difference, we genotyped SNPs for ancestry informative marks (AIMs). we
calculated population proportions based on these AIMs. Each subject has 4
population proportions. The sum of these 4 population proportions is equal
to 1. For example, for subject A, pop1 = 0.7, pop2=0.1, pop3=0.15, pop4=0.05
. These 4 population proportions provide ethnicity information. If we add 4
population proportions into the model, there will be a collinearlity problem
. What's the propriate way to add these 4 population proportions into the
model?
Thank you. | t*****w 发帖数: 254 | 2 我是从生物转统计的。
你的model 是
y=pop1+pop2+pop3
removal of any population can solve the collearity issue in your model. | p*****V 发帖数: 43 | 3 Thank you.
【在 t*****w 的大作中提到】 : 我是从生物转统计的。 : 你的model 是 : y=pop1+pop2+pop3 : removal of any population can solve the collearity issue in your model.
|
|