|
|
|
| Koadarray
revision history |
| The software
changes for each version of Koadarray are listed below
in reverse chronological order. Please feel free to contact
us if you have any questions about these
changes, if you have any suggestions for improvements
or if you would like to report a bug. |
|
V2.6.49
- August 2007
New features and changes:
1) Spot finding algorithms have been further improved.
This gives better spot finding efficiency in lower
quality images, specifically if there are few or faint
spots. Some default spot finding algorithm parameters
have been changed to match the improved algorithms.
2) The spot intensity threshold parameter has been
removed from the Options window, as this parameter is
now determined automatically by the improved spot
finding algorithms.
3) The spot shape threshold parameter has been
reversed so that it matches the other two spot
threshold parameters (i.e. a lower value is now more
stringent).
4) When opening single monochrome images the software
now checks if the image seems to be negative. If it
is, it shows error 13, then reverses the grey scale if
user affirms. If multiple images are being processed
the grey scale is reversed automatically.
5) A 33% setting has been added to the image zoom
drop-down box.
6) An additional error trap (warning 352) has been
added if a calculated or found block position finds no
spots within it.
7) A new menu item has been added to allow spot IDs to
be cleared.
8) Copyright notices have been changed from 2005 to
2006.
9) The number of analyses permitted in evaluation mode
has been increased to 30.
10) A note to inform evaluation users that an
extension utility is available has been added to the
Contact info and copy status window.
Bug fixes:
1) Dual files are now correctly put in the last used
files list and can be re-opened from there.
2) A bug that was stopping legacy Molecular Dynamics .ds
files being opened has been fixed.
3) A bug that caused an error in the evaluation status
in Korean versions of Windows has been fixed.
4) The number of image files permitted in the Multiple
image analysis window is now correctly limited to
three in evaluation mode.
5) A file format incompatibility in the GenePix data
output that was causing an error when loading into the
limma Bioconductor package has been resolved. The same
modification was also implemented in the Koadarray
output format, though this is not known to cause any
problems.
6) A problem with evaluation status in US date format
Windows systems has been fixed.
7) If a dual image TIFF file is opened in which the
two images are different sizes, the option is given to
open only the first or the second image, or attempt
overlay of both.
|
|
V2.5.30
- November 2005
New features and changes:
1) Spot finding algorithms have been significantly
improved. This gives better spot finding efficiency in
lower quality images, specifically if there are few or
faint spots, or if blocks are missing. Some default
spot finding algorithm parameters have been changed to
match the improved algorithms.
2) A facility to output data in GenePix (.gpr file)
format has been added. Details of the data and format
are given in the Help file. This facility is set from
the Options window.
3) The Options window has been reorganised into three
tabbed sections - Data output, Spot finding and
Automation. Each section has its parameters reset to
default values separately when the button is clicked.
4) The image opening system now checks TIFF image
description tags and file names to see if there is
text that specifies which channel is red and which is
green. If no such text is found then Ch.1 is set as
red and Ch.2 green. New menu items for zooming,
intensity scaling and switching the channel colours
have also been added.
5) The column headings for two channel data in
Koadarray data output have been changed so that the
channel label is a prefix to the heading (e.g. Ch.1
Sp_Mean) rather than in the line above it.
6) New data types have been added to the Koadarray
data output format (Block number, Spot X pixel
position, Spot Y pixel position and Spot diameter).
7) Some of the Koadarray data types have been renamed
to give a more structured naming system. They have
also been put in a more logical order in the list box
in the Add/remove spot data types window. The window
has also been modified slightly.
8) An ID file can now be opened from the command line
using the parameter "ID". The parameter for
opening an image file has been changed from from
"IF" to "IM".
9) The ignore warnings system has been changed so that
the warning message is now shown, but the word
"IGNORED" is added. The ignored warning
still has no effect on the analysis process.
10) The facility to save data in a 'blocks' format
instead of a 'columns' format has been removed.
11) The data file header/footer now shows average spot
diameter instead of average radius. Also an
unnecessary carriage return has been removed.
12) Error 51 ("Analysis may have failed")
now shows the Koada support email address.
13) A blank line is no longer added in the
analysis log after the last image processed. This
means that when the user clicks on "View
data" it automatically shows that last data file
without the user having to select a line in the log.
14) The automatic intensity scaling function has been
improved to make images clearer, especially when spots
are faint.
15) The toolbar buttons relating to image
analysis are now disabled until an image is opened.
16) The spot selection system has been modified
slightly so that a mouse click in one place after
selecting spots will deselect them.
17) To improve clarity the text used in notifying spot
finding processes has been changed to
"cycles" and the text in Options window -
Spot finding has been changed to "more
stringent".
Bug fixes:
1) A bug has been fixed where if a 16 bit TIFF is
opened then an 8 bit TIFF, if the mouse is in the
right half of the screen a subscript out of range
error results.
2) A bug has been fixed which could cause marginal
inaccuracy in the calculation of the background
standard deviation in cases where background level
around a spot varies widely.
3) A bug has been fixed in which the data file name
extension was not being reset to the default in the
Options window.
|
|
V2.4.57
- September 2005
New features and changes:
1) The number of permitted image analyses in
evaluation mode has been increased to 25.
Bug fixes:
1) A bug has been fixed that can cause an error in the
evaluation status check if the software is used in
evaluation mode on or after September 4 2005. |
|
V2.4.56
- March 2005
New features and changes:
1) The spot circles on screen are now displayed in a
colour indicating the confidence index, varying from
red (low confidence) to blue (high confidence), so
giving the user immediate feedback about the quality
of the array. These colours are only shown after spot
data have been calculated and the confidence index is
known. Circles are all shown in green if spots have
been found but data have not yet been calculated.
2) Multiple spots can now be selected and then moved
manually, or have their confidence indices set to
zero. This can be done by drawing a box around them
and/or clicking on them individually. Selected spots
are shown in orange.
3) To facilitate integration of Koadarray with other
laboratory systems, it is now possible for Koadarray
to accept command line parameters to automatically
open an image or pair of images, analyse them, then
close after the analysis is complete.
4) The facility to set the image zoom using a dropdown
box has been provided. This allows the user to select
one of several preset zoom settings, fit the image to
the current window height or width, or type in a zoom
setting. The maximum zoom is also now restricted to
200%.
5) The default values for "Spot shape confidence
threshold" and "Spot intensity
threshold" have been increased to 35% and 25%
from 30% and 20% respectively. This should enable a
broader range of problem images to be analysed
reliably.
6) The facility to save a copy of the analysed image
in JPEG format in addition to bitmap format has been
provided.
7) The installation and evaluation copy protection
systems have been changed to allow installation by
non-administrator users. However, administrator
privileges are still required to install the hardware
key and drivers. A maximum of 20 images is now
permitted for each evaluation user.
8) The spot finding algorithms have been improved to
make better allowance for spots that are substantially
larger than most in the array.
9) Support for "MM" byte order TIFF files
has been provided. This means they use 'big endian'/Motorola
byte order, in which the byte order is always from
most significant byte to least significant, for both
16-bit and 32-bit integers. Previously only Intel
("II") byte order TIFF files were supported.
10) Automatic intensity scaling is now performed on
analysed images before being saved as a bitmap or
JPEG.
11) The block finding/verification algorithms have
been modified to improve reliability if they are not
able to unambiguously find the top and bottom rows,
and also to prevent an occasional failure to reliably
find the top row of a block even when there are some
distinct spots in it.
12) The spot finding algorithms have been improved to
increase reliability if the top left block is a large
distance from the edge of the image.
13) The spot finding algorithms have been improved to
improve spot positioning in very long rows or columns
with no detectable spots, or few detectable spots in
the first row or column.
14) The spot finding algorithms have been improved to
reduce the occurrence of spurious spot perimeter
overlapping errors (error 357).
15) The currently open spot ID filename is now saved
in the output data file header/footer and is also
shown in the title bar of the main window.
16) The software version number is now shown as VX.X.X
instead of VX.X (Build X)
17) The software version number is now shown in the
title bar of the "Copy status" window.
18) Minor changes have been made to the spot ID file
opening code to check for header text if any
non-numerical characters are found up to the first tab
character in the first line, and also to trap empty
files.
19) Non-numerical inputs for numerical parameters in
the Options window are now trapped, giving error
number 209.
20) Warning 350 "used >50% of maximum
attempts" has been removed from the spot size and
spacing finding code as it was not providing any
useful feedback.
21) Warning 352 message has been simplified to
"Unable to find block".
22) The mouse pointer shown during spot finding and
calculation processes has been changed to arrow +
hourglass. Also the help button has been disabled
during spot finding and calculation.
23) Non-numerical inputs in any numerical input boxes
are now prevented.
24) The block and spot number fields are now
highlighted when the relevant text box is selected, so
enabling numbers to be changed more easily using the
keyboard.
25) Error handling in the event of memory limitations
has been improved, giving new error number 118
"Error creating image copy".
26) Error handling in file save operations has been
improved, giving new error number 119 "Error
saving image file".
27) The default number of blocks and spots has been
changed to X blocks = 4, Y blocks = 12, X spots = 24,
Y spots = 24.
28) Moved menu item "Save Data" to the
"File" menu. Changed menu item "Clear
Data" to "Clear Positions and Data".
Bug fixes:
1) A bug has been fixed in which the channel 2 volume,
mean and median values were not being set to zero if
this parameter was set in the Options window. The text
in the Options window was also changed to "Set
negative spot values to zero:"
2) A bug has been fixed where if an image is opened,
the image selection dialog box is then opened but
cancelled, then spot finding is attempted, the
software does not perform the analysis but gives an
"Analysis OK" message, then a
"Subscript out of range error" if spot
finding is attempted again.
3) A bug has been fixed which could occasionally cause
a "Run-time Error 52: Bad file name or
number" error when opening some images on some
PCs. Additional error trapping has also been
implemented.
4) A bug has been fixed which could cause an
additional "Subscript out of range" error in
the event of an error 107 on saving a bitmap of an
analysed image.
5) A bug has been fixed that causing the vertical
scroll bar not be shown if the image size was very
slightly larger than the available window space.
6) A bug has been fixed that would cause an endless
loop if the text in one of the block or spot number
fields was deleted and no text input.
7) A bug was fixed that was preventing a new spot
being located when a single spot circle was manually
moved to a detectable spot. This meant that the spot
confidence index was remaining the same as the
previous location. This very seldom occurs, because if
there is a detectable spot it should not be necessary
to manually move the circle. |
|
V2.31 -
September 2004
1) A bug has been fixed in the spot ID feature (V2.1)
in which the block sequence of some spot ID files
could be read incorrectly. This depended on the block
layout being used, but could result in spot ID labels
being allocated incorrectly to some blocks other than
the first and last. The bug affected only the block
sequence, and the sequence of spot IDs within each
block was unaffected. All other spot data types and
values, whether in the output data file or shown
on-screen, also remained correct. |
|
V2.3 -
July 2004
New features and changes:
1) If the Options parameter "Calculate position
and radius if spot not found" is set to false,
any data for unfound spots other than Spot Flag and
Spot ID will be output as "U".
Bug fixes:
1) A bug has been fixed where if the number of blocks
or spots was changed and then the 'Find' or 'Find and
calculate' button was clicked immediately, the changed
number of blocks or spots was not used. |
|
V2.2 -
June 2004
New features and changes:
1) The algorithm for re-aligning two colour images
that have been scanned slightly out of phase has been
improved for speed and range of misalignment.
2) Error handling has been added to the function to
enable the Koada Technology web site to be opened from
the Koadarray main window.
3) The software terms and conditions of use have been
modified to improve clarity, in the 'About' box, Help
file and installation package.
Bug fixes:
1) A bug has been fixed that could potentially cause
an error in reading TIFF tags in Molecular Dynamics
.GEL files. |
|
V2.1 -
April 2004
New features and changes:
1) The facility to open a tab-delimited text spot ID
file in association with each image file has been
implemented, whether analysing single or multiple
images. The ID information for each spot can then be
automatically saved in the Output data file and/or
viewed on screen by holding the cursor over a found
spot.
2) A new spot data type, "Standard deviation of
background pixel values" (Bgd_SD) has been added.
3) A new spot data type, "Spot flag" (Sp_Flag)
has been added to facilitate downstream analysis with
software such as GeneSpring. This flag can be
'present' (P), 'absent' (A) or 'marginal' (M) and is
calculated from the spot confidence value and two new
threshold values that are set in the Options window.
4) Multiple processing of pairs of monochrome images
is now supported.
5) When manually moving a block of spots, the whole
block is now shown moving. This makes it easier to
drop it in the right place, especially if the top left
spot is not visible.
6) The default minimum spot diameter has been changed
to 5 pixels. This will enable more reliable spot
finding if some spots are very small.
7) The default spot finding option "Use average
spot radius for all found spots" has been changed
to unchecked. This means that by default the actual
spot radius for each spot will be used, if it has been
found.
8) The facility to use any text as the data output
file headers for the two channels has been added (e.g.
"red" and "green" instead of
"Ch1" and "Ch2"). The text for
these is set in the Options window.
9) The layout of the Options window has been modified
to allow the Spot Flag thresholds and channel headers
to be set.
10) The order in which 2-colour data is saved has been
changed. The channel 1 and channel 2 data for each
data type are now saved next to each other, instead of
all the channel 1 data then all the channel 2 data.
11) The Main window has been changed so that the
image, pixel and spot data for the two channels are
shown in two boxes rather than one box with a
separator character.
12) The calculation of errors 353 and 354 (block
horizontal and vertical positions may be incorrect)
has been improved.
13) If the same spot ID file is required for all the
image files in multiple image analysis, it is now
possible to put just one ID file in the list box
rather than adding multiple copies of the same ID
file.
Bug fixes:
1) A bug has been fixed that meant was only channel 1
being used for spot finding when re-finding spot
positions after manual block adjustment. This meant
that the re-finding was sometimes less effective than
the initial spot finding, which always uses both
channels.
2) A bug has been fixed that was preventing multiple
files being added to the image or spot ID file lists
in multiple image analysis if the combined number of
characters in their filenames exceeded 260. |
|
V2.0
- October 2003
New features and changes:
1) Support for two colour channels has been
implemented. Single or dual-image files are now
automatically detected, opened, displayed and analysed
appropriately. In dual-image TIFF files, the first
image (channel 1) is assigned the colour red and
channel 2 is green. In Molecular Dynamics .DS file
pairs the reverse colours are assigned. When a TIFF
file containing two images is opened, dual sets of
image or spot information are shown on the toolbar and
status bar. The two channels can be hidden or shown
using buttons on the toolbar. Spot finding is carried
out on both channels simultaneously and the resulting
data are saved in a single data file, containing Ch.1
and Ch.2 labels as appropriate. If an artefact is
found in a spot in one channel, the artefact removal
process is also carried out on the other channel.
2) The facility has been added manually open two
separate monochrome 16 bit TIFF image files as a dual
colour image. This is done by selecting two files
instead of one in the file selection dialog box. The
first file selected will be channel 1 (red) and the
second one will be channel 2 (green). This facility is
not available when opening multiple files for
unattended analysis.
3) When opening two separate files as a dual
colour image, the output filename is now
filename1[filename2].ext instead of
[filename1]filename2.ext.
4) A new spot data type, "Background subtracted
spot median" (Sp_Med) has been added.
5) Two-colour images in which the two channels
are out of alignment by up to 10 pixels are now
overlaid and aligned automatically.
6) The spot confidence index calculation has been
changed so that it gives a more intuitive result
range. Now spots with indices above 40 or 50% can
usually be regarded as reliable.
7) The data file header or footer now contain some
measures of array quality, including the proportion of
spots found, background variation and mean spot
confidence.
8) The spot data displayed on the status bar when the
cursor is over a spot can now be configured to show
any two spot data types.
9) The Koadarray logo and program icon have been
changed to reflect the two-colour capability.
10) After spot finding and data calculation, a right
mouse click on a spot raises a menu that allows the
confidence index for that spot to be set to zero.
11) A new menu item has been added to allow
manual saving of spot data (Spot Finding.... Save
Data) which may be needed after any manual adjustment
of spots or blocks, or after setting spot confidence
values to zero.
12) Installation of this version will cause the 30-day
evaluation period to be set to day 1, irrespective of
when any previous version was installed or used.
13) The spot artefact exclusion stringency range has
been modified slightly. The default value of 3 now
gives better results across a wide range of images.
14) The default spot data types have been changed to
the spot/block numbers, spot mean and spot confidence
values. These are probably the most commonly required
data. Also the default is now NOT to use the average
spot radius for all found spots.
15) It is now possible to add several files at a time
to the Image file list window by using the CTRL or
SHIFT keys
16) When any changes to the Zoom, Hi or Lo fields, the
change is implemented when the focus is lost from that
field. The Apply button has therefore been removed.
17) The individual limits on number of blocks or spots
have been removed and replaced by a trap that prevents
the total number of spots exceeding 250000. The new
error is code 204.
18) The automatic intensity scaling has been improved
so that it is faster and works better with 2 colour
images.
19) A spot finding algorithm change (point 7 in V1.7
revision history) has been removed because it caused a
reduction in spot finding reliability in some cases.
20) If an image is opened that contains no intensity
scaling information then automatic intensity scaling
is called automatically.
21) The automatic intensity scaling button text has
been changed to Auto.
22) Image inversion (Neg button) is now enabled only
in monochrome images.
23) The text of error 105 has been changed to "RGB
or palette colour TIFF images not supported. Image
cannot be opened".
24) Error 102 has been changed so that if any TIFF IFD
(image file directory) contains no TIFF tags then a
non-fatal error is generated.
25) The add/remove spot data types window has
been enlarged.
26) The default has been set back to using average
spot radius, as this produces generally more reliable
results across a range of images.
27) A new error (code 115) "Only one or two
image files may be selected" has been added.
28) The text of File menu items and toolbar button
'tips' has been changed from "file" to
"image" (e.g. "Open File" to
"Open Image"), because an image can now be
two files.
29) The text of error 106 has been changed to changed
to "Invalid file type. Image cannot be
opened". This error is now also used to prevent
pairs of files other than 16-bit TIFFs being opened as
a dual colour image.
30) A new error trap has been added to prevent two
monochrome image files of different sizes being opened
as a dual colour image. This uses the existing error
code 104.
Bug fixes:
1) A combination of bugs has been fixed that could
cause anomalous ratio data to be generated in some
pairs of images. The anomalies appeared to be caused
by artefacts being removed from spots in one image and
not in the other, combined with the Sp_ArtX flag being
incorrectly set.
2) A bug has been fixed in the spot mean (Sp_Mean)
calculation. This caused some inaccuracy which was
apparent when comparing repeat analyses of some
images, and was primarily due to use of an integer for
storing spot radius rather than a floating point
number.
3) A bug has been fixed in the spot median (Sp_Med)
calculation, in which under some circumstances the
pixel values were not sorted by value prior selecting
the median value.
4) A bug has been fixed that can cause the
evaluation period check to fail with some Windows date
formats, including United States.
5) A bug has been fixed that caused the evaluation
period to incorrectly be reported as exceeded when
using a logon different to that used when Koadarray
was installed.
6) A bug has been fixed that caused the wrong data
file to be opened if the filename line was selected in
the Progress window and then View data was clicked.
7) A right mouse click to set spot confidence to
zero is only enabled after spot data have been
calculated.
8) A new error trap has been added (code 58
"Spot data from this image have been changed. Do
you want to save the changed data?") which is
triggered if the user tries to open new images, exit
or change spot/block numbers when any spot data have
been changed.
9) A bug has been fixed that improves the
accuracy of background calculations when spot spacing
is very close.
10) A bug has been fixed that was causing only
partial merging of the two channels prior to spot
finding. This bug was unlikely to cause any problems,
but fixing it will slightly improve spot finding
reliability. |
|
V1.71 -
July 2003
1) A bug has been fixed that caused incorrect data
files to be saved during multiple file analysis.
Specifically, each data file was being saved using the
name of its predecessor in the file sequence. |
|
V1.7 -
July 2003
1) The Sp_Mean data type has been changed so that it
is now background subtracted. This is more likely to
be useful and brings it into line with the Sp_Vol data
type. The name has also been changed from "Mean
of spot pixel values" to "Background
subtracted spot mean".
2) In order to support more fully automated analysis,
a facility has been added for running a program or
file after each image analysis. This facility may be
used for purposes such as automatically opening an
Excel file for data processing. The parameters for
this facility are defined in the Automation section of
the Options window.
3) A new option has been added to allow the output
location (for data files and bitmaps) to be selected
rather than using the same location as the image file.
4) A new dialog box has been added (accessed from
File…. Image properties in the Main window) showing
properties of the currently displayed image, including
TIFF tag details.
5) Support for ".tiff" and ".jpeg"
extensions has been added (in addition to ".tif"
and ".jpg").
6) The block search algorithms have been slightly
changed to improve reliability of finding misplaced
blocks.
7) The spot finding algorithms have been augmented to
improve reliability of spot finding in images with
inconsistent spot positions.
8) The "Modify" button in Options window and
title of resulting window have been changed to
"Add/remove".
9) The options file format has been changed to give
more flexibility.
10) The image and data file paths are now written to
the registry so that they can be read by external
programs.
11) Error traps have been added to options file open
and save routines. Three new error codes (111, 112,
113) have been added.
12) TIFF files using the "MM" reverse byte
order format are now trapped and an error reported
(error code 104).
13) A bug has been fixed that caused a "Type
mismatch" error when saving an options file.
14) A bug was fixed that was reducing spot finding
effectiveness if an image without a TIFF tag 33445
value (pixel data format) was opened after one with a
value for this tag. |
|
V1.6 -
April 2003
1) The calculation of the values for the data types
"Mean of spot pixel values" (Sp_Mean) and
"Standard deviation of spot pixel values" (Sp_SD)
has been changed so that rather than being calculated
before artefact exclusion, they are now calculated
after artefact exclusion.
2) A new data type "Spot artefact pixels
excluded" (Sp_ArtX) has been added, which shows
for each spot whether artefact pixels were excluded.
An asterisk(*) is also shown after the number in the
"Vol" box on the status bar when the cursor
is moved over a spot from which artefact pixels have
been excluded.
3) The stringency used by the artefact exclusion
algorithm can now be set in the Options window as
"Spot artefact exclusion stringency", with a
valid range of 1 to 10. An incorrect input will cause
error 226. |
|
V1.5 -
April 2003
1) Part of one of the spot finding algorithms has been
changed on order to improve reliability in images with
few detectable spots.
2) The text of the error 357 message has been changed
to "Spot perimeters are overlapping". |
|
V1.4 -
April 2003
1) A bug has been fixed that gave an error 104 on
opening Agilent TIFF image files. This was caused by
incorrect use of Molecular Dynamics TIFF tag number
33447 (MD_COLORTABLE) in the Agilent files. The tag
should have >4 values, but the Agilent files appear
to encode only one. Not only was the tag being used
incorrectly, it should only be used in MD files
conforming to the MD .GEL file specification, not
generic TIFF files.
2) A bug has been fixed that caused a "subscript
out of range" error if there was a file opening
error (e.g. error 100) during multiple file analysis. |
|
V1.3 -
April 2003
1) A bug has been fixed that was causing an error 104
to be shown in a small number of images due to TIFF
tag 279 sometimes being encoded before tag 273 in the
TIFF structure. |
|
V1.2 -
April 2003
1) A bug has been fixed that was causing an incorrect
TIFF data offset to be read in a small number of
images, so causing the image display to be 'wrapped'
mid-image. |
|
V1.1 -
March 2003
1) A bug has been fixed that was causing an
"overflow" error in one test image when
"Automatic intensity scaling" was used after
"Invert image grey scale".
2) The block finding algorithm has been changed so
that block searches start 1 spot spacing from the
block above/left. Previously it was 1 spot spacing + 1
spot radius.
3) A potential bug problem was fixed to prevent
"divide by zero" errors with some spot
position threshold values.
4) One of the spot finding algorithms has been changed
to make spot finding more reliable in arrays where x
spot spacing is compressed or expanded in part of a
block (i.e. spotter or scanner problems).
5) A bug has been fixed that was preventing the
background from being subtracted from spot volumes.
This bug was inadvertently introduced in beta version
0.99.
6) The errors/warnings text has been changed so that
several # symbols are shown in the final note about
whether the spot finding or data calculation had
errors or warnings. None are now shown in the
individual errors or warnings. This highlights the
critical issue that a given image may have failed. |
|
V1.0 -
March 2003
1) The system for configuring the data file format has
been changed so that the available spot data types can
be defined in any order and with any heading. This
involves a new "Modify spot data types"
window, accessible from the Options window, in which
data types can be added/removed or their order and
heading changed.
2) A bug has been fixed that was causing incorrect
evaluation status to be reported under Windows 98 and
Windows Me. |
|
|
