R/qtl version 1.10-27

[Karl Broman]

A new version of R/qtl (1.10-27) has been released at its web site
(http://www.rqtl.org).

See http://www.rqtl.org/STATUS.txt for a list of changes. The major
changes are listed below.

As always, I'd appreciate hearing any suggestions or comments or
complaints.

Karl Broman
kbr...@biostat.wisc.edu

------

Version 1.10, 1/11/2009:

Major changes:

Revised the way that the 'chr' argument is treated in functions such
as scanone, scantwo, etc., to give greater consitency. Numbers are
interpreted as character strings to be matched to the chromosome
names. Negative numbers and character strings that start with "-"
are interpreted as omitting the corresponding chromosomes, matched
by name. One may also use a logical vector (TRUE/FALSE), of the
same length as there are chromosomes, indicating which chromosomes
are to be considered. So if an object has three chromosomes named
"1", "3", "4", using chr=2 will result in an error, while chr=3 will
give the second chromosome (named "3").

Also revised the way that the 'ind' argument is treated in
subset.cross and plot.geno, in the case that the input cross
contains individual identifiers in the phenotype data. The 'ind'
argument can still be a logical vector, but otherwise we first seek
to match the values against individual identifiers. For identifiers
that are character strings, one may use "-" at the beginning of each
to indicate all individuals except those given.

Added an argument 'batchsize' to scanone and scantwo, so that in the
case that multiple phenotypes (or permutations) are to be run as a
batch (with method "hk" or "imp"), they can be run in smaller
batches (indicated by batchsize). This can speed things up quite a
bit in the case of a very large number of phenotypes (or
permutations).

Added two functions for the de novo construction of a genetic map.
formLinkageGroups uses pairwise marker linkage information
(calculated with est.rf) to partition markers into linkage groups.
orderMarkers uses a quick but not very good algorithm for ordering
the markers on a chromosome (minimizing the number of obligate
crossovers).

Added a function addcovarint, which is similar to addint, but adds
one QTL x covariate interaction at a time.