May be correlation of the trait you calculated is greater than 0.5.

Dear Dr. Simona Kralj-Fišer ,

Greetings, thank you for your question. We estimate the heritability in our work (plant or crops).

The heritability value more than 90-100% it is a good one as it will be inherited traits and low than 50% it will be controlled by environments and the genetic value will be low.

I think you may be have some mistakes in your estimated equation.

with best regards

khaled

We got r > 1. How is that possible???

I'm not clear on the formula you used - it should be a Genetic cov (trait1, trait2) / sqrt(genetic var (trait 1) * genetic var (trait 2))

There will be no information on the residual (environmental) covariance as traits 1 and 2 are measured in different individuals (males and females).

http://www.wildanimalmodels.org/tiki-index.php?page=bivariate%20models gives an example of a MCMChmm bivariate analysis, including code to calculate a genetic correlation.

Hey, thank you. I am familiar with this kind of genetic correlations, but we need to get inter-sex genetic correlation and it seems to me that the above would be incorrect to apply as we do not measure correlation between two traits on the same individuals....

With some software you can model this as two traits - males have a value for one of the traits and missing for the other and the other way round for females. Genetic relationships allow estimating the genetic covariance, but there is no information on residual covariance which can be fixed at a value which you then ignore. I don't know if MCMChmm can do this, but worth looking at.

Dear Simona

This situation may result from the small sample size in your study, and the non-random mating in the population, that may be given abnormal results.

Good Luck

Any new research on Spiritual Care related to Post Partum Hemorrhage

 In reply to Ken G Dodds: "With some software you can model this as two traits - males have a value for one of the traits and missing for the other and the other way round for females. Genetic relationships allow estimating the genetic covariance, but there is no information on residual covariance which can be fixed at a value which you then ignore. I don't know if MCMChmm can do this, but worth looking at.

Inter-sex genetic correlation more than 1?. Available from: https://www.researchgate.net/post/Inter-sex_genetic_correlation_more_than_1 [accessed Jul 26, 2017]."

Thanks for this suggestion. We designed our matrix as proposed and run the prior and code: 

priorLar2

I'm not familiar with fitting models with MCMCglmm so I might be a bit off track here. It seems that your solutions are quite far away from your priors (e.g. prior genetic vars of 0.5, post.means of ~0.023). Perhaps using better priors would help, but in particular you should try a prior with a highish correlation, to make sure that converges to the same result.

I presume there are sufficient relationships between males and females to allow the genetic covariance to be influenced by the data.

The heritabilities are surprisingly similar ( 0.548 for both sexes).

Do you have access to a maximum likelihood - based (REML) software to see if that gives the same result?

Thanks Petri. Should run it this way.

Best,

S