This page contains a version/revision history to modifications and
additions to WHAMM and to the iHS calculator.
Modified use of C to C++ libraries for compiling purposes
(Thanks to Sébastien Letort for the suggestion)
Bug fixed where R optimization routine was referencing previous program name
(hmm_HZmapper.pl not WHAMM.pl).
released v1.0 of iHS Calculator
modified WHAMM to perfomr iHS prep and tabulations (--iHS-prep,
--iHS-scan, --iHS-subscan, --iHS-std).
added a --flip method to construct stand swapped .phased data files and associated
.bim files. [not currently implemented with .tped files].
added a --chr option for used with --iHS option to calculate
scores on specific chromsomes.
updated .R regression machinery to allow bug fixes
renamed the machinery "WHAMM" from hmm_HZmapper.pl.
modified default file output from "hzmap_out" to "whamm"
constructed my very first package (.tar.gz)
changed directory paths for NSP perl module and .R routines to
package path directory.
added Cwd.pm module for correct path lookup
added some error checking for .R routines before initialization
Fixed static path configuration for
hmm_hz_sim_optim_happrobs_v2.R call to a dynamic location
Added the genotype/heterozygosity framework. added --gt option,
default is to assume phased data model. changed some file preparation
--phaseloaderfile to --ploaderfile (Assumes .tped
--hapinfo to --infofile
--calc-happrobs to --calc-probs
Added a speedup/disk space efficiency measure to the regression
options --logistic and --linear: only kept samples are output to the
transposed .hzdose file fed into R
bug fix on get_gtdatafreqs line 2995
completely revamped the structure of the .R routine to perform
regression. Reads in data chunks at a time and performs regression,
which can be changed to allow for memory issues if the data set is
big (i.e., --make-overlapfile options).
modified the .R routine to accept covariates (first order)
added --covar option to the mapper, allowing the specification
of covariates into the regression
changed the functionality of the regression options (--linear
and --logistic); now outputs transposed .hzdose files for use with
the new .R routine (above)
added --make-overlapfile option which allows one to essentially
merge two data sets with different marker map information
fixed an error when no arguments are present on command line
(--estimate and $ARGV)
added --linkage option, --dosefile option, changed --regression
option to --linear and added a --logistic option.
bug corrected on --exttest l option where values != 0.5 weren't
obtaining the proper tail angle
added .hzlink file format
modified hmm_hz_dose_reg.R routine to utilize logistic and
modified --linear/logistic (formerly --regression) options to
process phenotype and keeplists and output processed files sent
directly to R (rather than calling R and operating directly on user
specified files -- useful for error control)
Encapsulated existing batch of scripts into a single perl
executable, which implements all of the basic features of HZ mapping
using phased data.
This document last modified Monday, 02-Apr-2012 17:35:45 EDT