|
Post by mandar on Jun 19, 2016 2:55:48 GMT
Hello, It may be a naive question but is the normality of the phenotype necessary for the stability estimate of heritability? (The reason for asking this question is ; I estimated heritability on bunch of phenotypes; grep results shown below with SE of the estimate. Plots for phenotype distribution attached) After looking at heritability estimates such as below, for set of 7 serial measurements of the same phenotype; do you have any suggestion about why it may be suddenly dropping heritability estimate to zero? PS. Serial measures are for glucose from Oral glucose tolerance test over 2 hrs. 1st at baseline 2 nd to last after 75 gm of oral glucose at different time intervals. V(G)/Vp 0.468111 0.300732 V(G)/Vp 0.000001 0.244915 V(G)/Vp 0.000001 0.222169 V(G)/Vp 0.000001 0.206394 V(G)/Vp 0.000001 0.260440 V(G)/Vp 0.052010 0.213405 Any comments or insights from experienced eyes will be helpful. Thank you -Mandar Attachments:
|
|
|
Post by Jian Yang on Jun 20, 2016 2:17:01 GMT
|
|
|
Post by mandar on Jun 20, 2016 3:45:12 GMT
Gotcha, Thank you for your response. My current sample size is 450. For all of my other phenotypes (other than above 7) SE ranges from 0.13 to 0.30. In any of these results (below), is there any "salvagable" information? Anything that can be used with "fair" certainty? OR when the sample size becomes lower, the estimates are just too unstable to be of any use.
V(G)/Vp 0.073751 0.179132 V(G)/Vp 0.000001 0.243429 V(G)/Vp 0.000001 0.251345 V(G)/Vp 0.003531 0.139100 V(G)/Vp 0.172436 0.219167 V(G)/Vp 0.495361 0.269088 V(G)/Vp 0.381265 0.238674
V(G)/Vp 0.827427 0.294171 V(G)/Vp 0.208101 0.237517 V(G)/Vp 0.172173 0.228851 V(G)/Vp 0.176853 0.229731 V(G)/Vp 0.091390 0.218783 V(G)/Vp 0.154524 0.206339 V(G)/Vp 0.158988 0.184152 Thank you -Mandar
|
|