Hi David,
The VFUNCTION procedure may be of help (https://genstat.kb.vsni.co.uk/knowledge-base/vfunctio/). This procedure can be used to calculate functions of variance components from a REML analysis, and provides an approximate standard error.
Genstat also has a VHERITABILITY procedure (https://genstat.kb.vsni.co.uk/knowledge-base/vherita) but this calculates broad-sense heritability.
Kind regards, Vanessa
-----Original Message-----
From: GENSTAT-Request <[log in to unmask]> On Behalf Of David Baird
Sent: Wednesday, 24 August 2022 12:14 pm
To: [log in to unmask]
Subject: Re: Getting an estimate of SE for narrow sense heritability
Dear David,
I would imagine the best way to get a model free estimate of this would be to bootstrap the analysis.
Take a random sample with replacement of the 421 horses and look at the variation of the h2 estimate.
You would need to recalculate the pedigree A inverse matrix for the new sample, although I'm I assume you would have to treat replicated horses as full sibs which may bias your calculations in some manner.
This is a problem in bootstrapping data with correlation models.
You can use the BOOTSTRAP procedure for this but that requires you write a RESAMPLE procedure encapsulating your analysis. The BOOTSTRAP help has information on how to write the RESAMPLE procedure.
Regards, David.
______________________________________________
Dr David Baird Statistical Consultant and Genstat Developer
VSN (NZ) Limited ([log in to unmask])
79 Ravensdale Rise, Westmorland , Christchurch 8025, New Zealand Cell phone +64 21 1160803
-----Original Message-----
From: GENSTAT-Request <[log in to unmask]> On Behalf Of David S Gardner
Sent: Tuesday, 23 August 2022 8:48 am
To: [log in to unmask]
Subject: Getting and estimate of SE for narrow sense heritability
Genstat community,
With help from others i have managed to run for the first time the above for 421 horses and actually managed to get an h2. i understand that h2 is a ratio of additive genetic as proportion of total, but surely there should be an estimate of error (SE) or 95% CI to go alonside the h2. Just wondering if i should be a maths geek to work it out from log2liklihood or whether there was an easy way of estimating it from my phenotypic outcome??
cheers
david
########################################################################
To unsubscribe from the GENSTAT list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=GENSTAT&A=1
This message was issued to members of www.jiscmail.ac.uk/GENSTAT, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
########################################################################
To unsubscribe from the GENSTAT list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=GENSTAT&A=1
This message was issued to members of www.jiscmail.ac.uk/GENSTAT, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
########################################################################
To unsubscribe from the GENSTAT list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=GENSTAT&A=1
This message was issued to members of www.jiscmail.ac.uk/GENSTAT, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
|