MapQTL 6 Release Notes
======================

=== 4 June 2013 ===
Maintenance release of MapQTL 6
  (Build: 4jun13.4feb11)

Minor internal adjustments.
Improved installer.


=== 12 March 2013 ===
Maintenance release of MapQTL 6
  (Build: 12mar13.4feb11)

An error was detected in the reported means and additive effects under
interval and MQM mapping of population type BCBxFy:
- mu_A contained the results of mu_B and mu_B those of mu_A;
- as a result of this, the additive effect contained minus the correct
  additive effect.
The error is corrected.
The means were reported correctly under the Kruskal-Wallis analysis.


=== 11 January 2011 ===
Maintenance release of MapQTL 6
  (Build: 11jan11.9feb10)

Additions:

The option 'Clear All Trait Nodes' is added to the
Edit-menu; it will clear the selections for analysis of all
traits of all populations.

A pop-up menu is added to select or de-select for analysis
all (numerical) traits of a population or all commmon
traits. The menu is activated by right-clicking on a
population node or the Common traits node, or by pressing
the space bar when a population node or the Common traits
node is selected.


=== 17 February 2010 ===
Maintenance release of MapQTL 6
  (Build: 17feb10.9feb10)

Minor internal adjustments.


=== 6 May 2009 ===
Final release of MapQTL 6
  (Build: 6may09.10mar09)

Changes:

In the Kruskal-Wallis analysis of population type CP under
the classification type (hh,k-) the results of class k-
were presented in the columns of class ll; this is
corrected.

The minimum number of permutations is reduced to 10 for
timing purposes.

The permutation test in the presence of design cofactors
and/or covariates was done in a theoretically incorrect
way. Now the permutation test is done on the data
residuals obtained by subtracting the effects of cofactors
and/or covariates (as estimated under the interval mapping
null hypothesis) from the trait values.


=== 17 April 2009 ===
Pre-release of MapQTL 6
  (Build: 17apr09.10mar09)

Changes:

The new manual is included as pdf document, available
under the Help-menu.

The pop-up menu for special selection of traits for
analysis is improved.

The workings of the "Common traits" node are improved.

Navigating in the results by clicking on group nodes
in the session tree now also applies to the 
"Results (cont'd)" tabsheet.

The permutation test in a combined analysis now produces 
a "Genome wide" node as a child of the Combined node.

The additive effect of cofactors in a BC1 was given a
factor 2 too small, now it is given correctly.


=== 10 March 2009 ===
Pre-release of MapQTL 6
  (Build: 10mar09.10mar09)

Automatic cofactor selection for traits combined over
populations is available.

This is the nearly final version of the program.
The rewriting of the manual is the only remaining task.
 

=== 13 February 2009 ===
Pre-release of MapQTL 6
  (Build: 13feb09.5sep08)

The permutation test for traits combined over populations
is available.


=== 20 January 2009 ===
Pre-release of MapQTL 6
  (Build: 20jan09.5sep08)

Populations can be combined in interval mapping, MQM and
restricted MQM mapping analysis. First the individual
populations are analysed, subsequently the combined LOD
score (or deviance) is calculated. Charts can be made
showing the combined LODs as well as the LODs and other
parameters of the separate analysis results.

(Automatic cofactor selection and permutation test for
traits combined over populations will follow later.)


=== 4 December 2008 ===
Pre-release of MapQTL 6
  (Build: 4dec08.5sep08)

Covariates and simple experimental design cofactors
(e.g. blocking) can be used with interval mapping,
(r)MQM mapping, automatic marker cofactor selection and
the permutation test, now both with the Haley & Knott
regression approximation and with maximum likelihood as
QTL analysis algorithm.


=== 26 November 2008 ===
Pre-release of MapQTL 6
  (Build: 26nov08.5sep08)

The results for the covariates and design cofactors are
now shown in a table. The results for the marker cofactors
as used to be shown in the cofactor monitor file are added
to this table, this cofactor monitor file is not used
anymore. The table is in the extra tabsheet called
"Results (cont'd)".

The preceding two releases contain errors in analyses
using covariates or design cofactors; these are corrected.


=== 14 November 2008 ===
Pre-release of MapQTL 6
  (Build: 14nov08.5sep08)

Covariates and simple experimental design cofactors
(e.g. blocking) can be used with interval mapping,
(r)MQM mapping, automatic marker cofactor selection and
the permutation test, but ONLY if the Haley & Knott
regression approximation is chosen as algorithm.
The results for the covariates and design cofactors
are not yet shown in a table (but will be later).


=== 11 November 2008 ===
Pre-release of MapQTL 6
  (Build: 11nov08.5sep08)

Covariates and simple experimental design cofactors
(e.g. blocking) can be used with interval mapping,
automatic marker cofactor selection and the permutation
test, but ONLY if the Haley & Knott regression
approximation is chosen as algorithm.


=== 16 October 2008 ===
Pre-release of MapQTL 6
  (Build: 16oct08.5sep08)

This release is not final. There are several features
that still need to be added, such as use of covariates
and simple experimental design cofactors. Also the manual
needs to be adapted to this version and printed.


Changes with respect to MapQTL 5:

= The code behind the user interface was overhauled to
  give an even more stable program. The interface itself
  hasn't changed much.

= The Haley & Knott regression approximation to maximum
  likelihood interval and MQM mapping was added.
  Advantages of the regression approach are a magnitude
  higher speed and reduced use of RAM memory, while the
  quality of the approximation is generally quite good.

= Three population types that can be analysed were added:
  DH2:
    doubled haploid population drawn from an F2,
    one doubled gamete from one F2 plant
  BCpxFy:
    advanced backcross inbred line family; starting from
    the BC1 repeatedly backcrossing to the same parent (as
    used for the BC1) of each individual resulting in a
    single offspring per individual, followed by selfing
    with single seed descent; the backcross parent p and
    the generations x and y must be specified:  p = A  or
    B,  x is the number of backcrosses including the one
    for creating the BC1:  1 <= x <= 99,  y is the number
    of selfings:  0 <= y <= 99,  BCa1F0 is equivalent to
    BC1
  IMxFy:
    advanced intermated inbred line family; starting from
    the F2 repeatedly random intermating (preferably chain
    crossing) the individuals resulting in a single
    offspring per individual, followed by selfing with
    single seed descent; the generations x and y must be
    specified:  x  is the number of intermatings including
    the two for creating the F2:  2 <= x <= 99,  y  is the
    number of selfings:  0 <= y <= 99,  IM2F0 is equivalent
    to F2, IM2Fy is equivalent to RIx  with  x = 2 + y


MapQTL 5 Release notes
======================


=== 23 July 2008 ===
Maintenance release of MapQTL 5
  (Build: 23jul08.7apr08)

= When no project is open, pressing the OK button on the
  Environment and the Analysis Options dialogs will save
  the options as default options that will apply to all
  future projects (similar to the 'Save as default' button).
                 
= The About box was renewed.


=== 9 April 2008 ===
Maintenance release of MapQTL 5
  (Build: 9apr08.7apr08)

Changes:
= A few minor (internal) improvements were made.


 === 4 April 2008 ===
Maintenance release of MapQTL 5
  (Build: 4apr08.11mar08)

Changes:
= An error was corrected in the (r)MQM calculations for the
  RIx population type. The error affected the outcome in
  situations with a high level of missing marker
  information, for instance when all markers on a linkage
  group are scored in a dominant fashion. In such cases the
  estimated genotype means and additive effect would be
  different, the LOD score and the explained variance were
  not very much affected.

= Rounding errors could cause the LOD score in interval
  mapping to be negative, with extremely small values like
  -10^(-13). This would affect the permutation test
  resulting in an error message 'error in
  permutation_test()'. Both in interval mapping and MQM
  mapping such negative values are now truncated to 0.0.

= Also due to rounding errors interval mapping and MQM
  mapping could occasionally fail to compute the results
  for the last marker on a linkage group.
  This is corrected.


=== 7 May 2007 ===
Maintenance release of MapQTL 5
  (Build: 7may07.4oct06)

Changes:
= MapQTL 5 was enabled to run under Windows Vista.
  For installation under Windows Vista, please see
  the 'Installation ReadMe' file.


=== 4 October 2005 ===
Release of MapQTL 5 (Build: 4oct05.1jun05)

Changes:
= If under (Restricted) MQM Mapping no null hypothesis (H0)
  could be calculated on an entire linkage group (due to
  singularity or perfect fit) a runtime error occurred;
  this is corrected.


=== 15 March 2005 ===
Release of MapQTL 5 (Build: 15mar05.29dec04)

Changes:
= When opening a project the linkage group names for each
  session are read from the Session Info tabsheet, however
  only the first 200 characters from the appropriate line
  were read whereas with longer group names and many
  groups the line would be much longer, which resulted in
  error messages like 'Error reading chart options ...'.
  This is corrected and now the group names are printed
  on the Session Info tabsheet over multiple lines (if
  necessary) and all of them are read.


=== 18 August 2004 ===
Release of MapQTL 5 (Build: 18aug04.2feb04)

Changes:
= The GIC (genotypic information coefficient) for the
  CP population was not output completely, only for the
  first parent (GIC_1). This release displays all GIC
  values for the CP population: GIC_1, GIC_2, GIC_m.


=== 20 July 2004 ===
Release of MapQTL 5 (Build: 20jul04.2feb04)

Changes:

= The GIC (genotypic information coefficient) for the
  RIx population type was calculated as for an F2, as
  shown in the manual. However the GIC for an RIx should
  be generation dependent, which is corrected. The right
  GIC formula for RIx is that of the F2 in the manual
  where the factor 2 before the summation is replaced by
  1/(2*p), with p the Mendelian expected frequency of the
  A genotype for the given generation x in RIx; e.g.,
  for RI2 (= F2) p=0.25, thus the factor 1/(2*p) is 2;
  for the 'infinity' generation, p=0.5, thus 1/(2*p)=1.
  This change results in higher GIC values (for RIx only)
  especially in between marker positions.

= The percentage progress is shown on the Windows task
  bar next to the program name.

= MapQTL could not deal correctly with situations
  where the Windows regional settings did not use the
  point as the decimal separator. Now MapQTL is forced
  to use the point as the decimal separator (i.e. within
  MapQTL only), so that everything functions correctly
  if there is a different decimal separator than a point,
  with the exception that sorting columns of numbers with
  decimal parts is done as if the numbers are text (this
  is due to an internal problem with the used ActiveX
  component for the tables).


=== 2 March 2004 ===
Release of MapQTL 5 (Build: 2mar04.2feb04)

Changes with respect to the pre-release:

= The user manual is finished.

= The Results tabsheet is filled only after the
  calculations (KW, IM, MQM, rMQM) are done for all
  linkage groups; this used to be done concurrently with
  the calculations which made everything rather slow.

= The Calculate button hint (shown when the mouse pointer
  is over the button) will tell what is required to
  enable the button in case it is not enabled.

= The chart axis titles were not drawn with the chart font
  indicated on the environment options; this is corrected.

= The preset default mapping step size is now 1.0 cM.

= The Session Info tabsheet lists the used statistic:
  LOD or Deviance. 


=== 4 Feb 2004 ===
Pre-release of MapQTL 5 (Build: 6jan04.2feb04)

This release is nearly final. Only the manual needs
to be adapted to this version and printed.


Changes with respect to MapQTL 4.0:

= The user interface was completely revised to give
  you more ease of use, better QTL charts and improved
  exportability:
  - multiple populations and maps in a single project;
  - analyses stored in sessions for (re-)inspection;
  - much improved QTL charts with many options;
  - data in user adjustable and exportable tables;
  - print preview.

= Computationally only the interval mapping of RIx
  populations was changed; a speed-up is achieved by
  setting an analysis option to use an approximation;
  the alternative is the original (slower) method;
  especially the permutation test of RIx benefits from
  the speed-up.

= The genotypic information coefficient is added.

= The license file system is overhauled:
  - when there is an evaluation or temporary license, a
    (paid up) license file can be installed using a
    Help-menu option;
  - license information can be viewed using a Help-menu
    option;
  - the licensed organisation is shown in the About-box.

= Options (including the list of last used projects) are
  stored in the 'MapQTL5' directory within 'My Documents'.
  If this does not exist it is created.