|
Date |
Version |
Changes |
|
8/29/08 |
2.6.1 |
Very minor cosmetic improvements affecting pedigree handling library and pedstats.
|
|
10/17/07 |
2.6.0 |
Upgrade to the core pedigree libraries to support compressed input files (except on Linux),
base-pair allele labels (A, C, T, G) and on the fly merging of input files.
|
|
01/12/05 |
2.5.1 |
Fixes to empirical p-value to resolve unexpected problems with version 2.5.0.
|
|
01/12/05 |
2.5.0 |
Switched to Mersenne random number generator, which produces greater randomness.
Smarted handling of allele numbers, avoids cluttering output with alleles that
are not present in the pedigree.
Packaged with a much improved version of Pedstats.
|
|
01/12/05 |
2.4.6 |
Fix to allele labeling bug in 2.4.4.
|
|
12/15/04 |
2.4.4 |
Many useability improvements to Pedstats. Fix to enable
comparison of maternal and paternally transmitted alleles
at the population level (-at -ot options).
|
|
09/17/03 |
2.4.3 |
Prelude fix to handle output IBD output files from Simwalk2 versions 2.86
and above.
|
|
05/01/03 |
2.4.2 |
New version of Pedstats optionally produces PDF output for checking trait
distributions (--pdf command line option).
More individuals are included in association tests when the total association
(-at) and imprinting options (-o*)are used together.
Activated --deviates in UNIX distributions.
Permutation tests now report empirical thresholds for .05, .01 and .001 significance
after multiplicity adjustment.
Original allele labels are now included in the regress.tbl when multi-allelic
tests are carried out.
|
|
12/12/02 |
2.4.0 |
Added --p-values option to print out all p-values, including those of
non-signifant tests.
Added --deviates option for data that is adjusted for the population
mean before analysis.
Added Fletcher-Reeves algorithm as an option for likelihood maximization.
Monks and Rabinowitz tests now assume test statistic follows a normal distribution,
rather than a t-distribution.
Fixed a bug in Windows version of pedstats that caused missing values to be
treated as zero.
|
|
17/03/02 |
2.3.0 |
Monte-Carlo calculation now provides global p-value, that adjusts
for multiple linked markers in a region.
Stopped QTDT from skipping individuals with missing covariate
values when covariates are disabled with -c- command line option.
|
|
25/01/01 |
2.2.1 |
Stopped QTDT crashing in multi-allelic test. Stopped QTDT from
trying to analyze discrete traits using variance components (need
a multivariate integral to do that!) |
|
18/01/01 |
2.2.0 |
Added support for discrete traits, choice of function for scoring
allelic transmission, a multi-allelic test, and optional inclusion
/ exclusion of founder phenotypes. |
|
28/03/00 |
2.1.11 |
Added support for twins and for Genehunter2 IBD calculations
(on a trial basis). |
|
09/03/00 |
2.1.9 |
Improved memory management (for C++ boffins out there, we switched
to virtual destructors). |
|
21/02/00 |
2.1.7 |
Transmission scores permuted correctly in entire family unit,
even when the requested tests assume independent probands (for
example, Rabinowitz test or simple linear regression models).
However, note that when multiple offspring per family are used,
it is better to use a model that allows for this (such as the
Monks' model or a variance components model). |
|
15/02/00 |
2.1.5 |
Provides more informative description of Rabinowitz' and Monks'
tests. In the Monks' test correctly handles the first individual
in the first pedigree. |
|
27/01/00 |
2.1.3 |
2.1.1 didn't actually fix the bug. This one might? |
|
26/01/00 |
2.1.1 |
Minor bug fix. Problem in allele counting when performing simple
inheritance check and interfacing with SimWalk2. |
|
18/01/00 |
2.1 |
First version available for public distribution. |
