| news | R Documentation |
Since this is the first documented NEWS release on the wgaim package it contains new features that have been included over several versions. In contrast the documented bug fixes are only recent.
The package now includes a PDF reference manual that is accessible by
navigating to the "doc" directory of the package. This can
be found on any operating system using the command
> system.file("doc", package = "wgaim")
The reference manual contains WGAIM theory and two thorough examples that show the features of the package. It also contains a "casual walk through" the package providing the user with a series of 5 steps to a successful wgaim analysis.
The package now includes three fully documented phenotypic and genotypic data sets for users to explore. Two of these three have been used in the manual and scripts that follow the examples in the manual are available under the "doc" directory of the package.
The package now provides very efficient whole genome QTL analysis of high
dimensional genetic marker data. All genetic marker data is passed
into wgaim.asreml() through the "intervalObj"
argument. Merging of genotypic and phenotypic data occurs within
wgaim.asreml().
wgaim.asreml() has several new arguments related
to selection of QTL. The "gen.type" argument allows the user to
choose a whole genome marker analysis or whole genome mid-point
interval analysis from Verbyla et. al (2007). The "method" argument gives you the choice of placing
QTL in the fixed part of the linear mixed model as in Verbyla et.al
(2007) or the random part of model as in Verbyla et. al
(2012). Finally, the "selection" argument allows you to choose whether QTL selection
is based on whole genome interval outlier statistics or a two stage process of
using chromosome outlier statistics and then interval outlier
statistics.
A "breakout" argument is now also provided which allows
the user to breakout of the forward selection algorithm at any
stage. The current model along with any calculated QTL components are
all available for inspection.
All linkage map plotting functions can be subsetted by predefined distances. This includes a list of distances as long as the number of linkage groups plotted.
cross2int() now accepts R/qtl objects with cross type
"bc","dh","rilsef".
Fixed an issue with the internal function fix.map() that
allowed some co-located sets of markers to appear in the final
reduced linkage map.
Fixed a long standing scoping issue with different versions of ASReml-R.
Fixed an elusive problem that causes wgaim models to increase the size
of your .RData upon saving. This is actually an inherent problem with
using model formula within a function a returning the subsequent
model. There is now a function at the very tail of
wgaim.asreml() that quietly destroys the useless environments
that these formula contain.
Fixed bug that caused wgaim.asreml() to crash when no QTL
were found.
Fixed bug that caused summary.wgaim() to crash when one
QTL was found using method = "random".