bqtl (version 1.0-3)

summary.adj: Summarize Laplace approximations

Description

The linear approximations of swap are much improved by the use a Laplace approximations for loci that are not markers. This function combines the results of a call like bqtl(y~configs(swap.obj),...) with the data in swap.obj to provide improved posteriors, et cetera

Usage

summary.adj(adj.obj, n.loc, coef.znames, mode.names=c("add", "dom"), imp.denom=NULL, swap.obj=NULL)

Arguments

adj.obj

Typically, this is the result of a call like bqtl(y~configs(swap.obj),...)

coef.znames

map.names for the sample space

mode.names

NULL except for "F2", in which case it is uusally c("add","dom")

imp.denom

Optional, and only used when some sampling scheme other than the default MCMC generates adj.obj

swap.obj

The result of a call to swap

Value

• A list with components
• adj
  This multiplier adjusts the posterior odds for k vs k-1 gene models
• var
  An estimate of the variance of adj
• coef
  Posterior means of coefficients
• loc
  Marginal Posterior for location for k gene model
• hk.ratio.mean
  argh! I need to look this up

Details

There are a lot of details. This sections nneds to be revised to reflect them.

References

Berry C.C. (1998) Computationally Efficient Bayesian QTL Mapping in Experimental Crosses. ASA Proceedings of the Biometrics Section, 164-169.