Revision history of Deconvoluter program.

Current version is 1.10 (2-26-05).

Earliest supported version is 1.10. If you are encountering problems, ensure that your current version is this version or higher.

Deconvoluter Version 1.10, 2/26/05
This represents a major maintenance release.
• New Feature: The Score Conversion Utility can now accommodate scoring systems that use continuous range scores (i.e. not discrete numerals). The Utility linearly converts the scores to a range of -2 to 2, with equivocal/uninterpretable scores set to 0. This range can be adjusted to something else, as long as the absolute values of the 100% negative and 100% positive values equal each other. Please refer to this page for details on how this conversion is accomplished.
• Modification: When a user clicks on "Reset", it no longer deselects the runtime option.
• Modification: The Deconvoluter will now make the following checks before running:
  • If the user attempts to run the Deconvoluter before resetting, the Deconvoluter will halt and request the user to either click on "Reset" or manually delete the output worksheets.
  • If the "Convert scores quantitatively" option is selected, the Deconvoluter will halt if the score key is not fully populated. The Deconvoluter will also warn the user that the post-conversion scores would not be TreeView compatible if equivocal/uninterpretable is set to a value other than zero, or if the absolute values of the 100% negative and 100% positive values do not equal each other.
  • If the user attempts to run the Deconvoluter without specifying either the scoring worksheet or the lookup file, the Deconvoluter will halt and request that the user populate those fields.
• Modification: The default scores for K-M output was changed from: negative = 0, equivocal = (blank), weak positive = 2, strong positive = 3 to negative = -2, equivocal = (boank), weak positive = 1, strong positive = 2.
• Bug Fix: An obscure bug was occuring where, if the user had selected K-M output or manual operation, the score conversion process was not selecting the data properly, resulting in non-conversion of the first row of data. This has now been fixed. Note: this bug did NOT affect PCL output in earlier versions.

Deconvoluter Version 1.08, 8/20/03
• Modification: Some minor changes were made to the user interface, mainly involving the score key. "Old Scores" and "New Scores" were clarified to mean "Before Deconvolution" and "After". This does not, however, affect the operation of the Score Conversion Utility, if operated separately as a manual process. Furthermore, the default score key now has the TreeView heatmap color representations located adjacent to their corresponding scores.

Deconvoluter Version 1.07, 4/1/03
• New Feature: Deconvoluter's Score Conversion Utility now supports the simultaneous conversion of up to 999 scores.
• Modification: The Score Conversion Utility can now be turned off by simply leaving all score fields blank. A pop-up message will ask to confirm this if the Deconvoluter is run in the default automatic mode or if the Score Conversion Utility is run separately.

Deconvoluter Version 1.06, 6/1/02
• Bug Fix: There was a bug that causes the Deconvoluter to navigate to the incorrect column if the number of columns in the sectors were different than that of the very first sector. Thanks to Brian Ring of Applied Genomics for pointing out this bug. This bug has now been fixed.

Deconvoluter Version 1.05a, 4/26/02
• Bug Fix: The Deconvoluter was not activating the "Output to cluster format" checkbox option when manual operation mode was selected. This has now been fixed.

Deconvoluter Version 1.05, 4/22/02
• Bug Fix: There was a subtle bug in how Deconvoluter handled different sector sizes in the master worksheet of the raw scoring workbooks. The TMA-Deconvoluter had been allocating memory incorrectly for TMAs which had the last sector be smaller than the other sectors, resulting in "Subscript out of range" errors. This has now been fixed.
• New feature: Deconvoluter now prompts you if you want to input your lookup file, after you generate your file list. If you select "yes", it will then ask for your lookup file and place the file name in the appropriate field, saving you the hassle of typing it in.

Deconvoluter Version 1.04, 4/7/02
• New feature: Deconvoluter now checks your lookup file to determine whether the list of unique case identifiers is correct. This will alert the user to any offset in the case numbering.
• Modification: Minor changes made to the "Control" worksheet (main screen) of the Deconvoluter.

Deconvoluter Version 1.03, 3/15/02
• Bug Fix: Deconvoluter had always assumed that the dimensions of all sectors would be equal to or less than the dimensions of Sector 1. As a result, cases existing outside the dimensions of Sector 1 in the other sectors were being omitted from the output file. This has now been fixed, and Deconvoluter should support TMA layouts where the dimensions of each sector is independent of any of the other sectors.
• Modification: Added an option to turn off warning messages for missing Bliss code/slice information. This is useful for people who haven't yet or don't wish to establish operation with Stainfinder at this time.

Deconvoluter Version 1.02, 3/5/02
• Modification: Made the "Set Working Dir" more informative for the user.

Deconvoluter Version 1.01, 2/8/02

• New feature: enhanced file list generation. Now, instead of prompting the user whether any of all the XLS files found in the current working directory are to be added to the file list, the user gets an Open File dialog box, whereupon selecting the desired files and clicking "Open" will result in the addition of those files into the file list.
• New feature: "Set Working Dir" button. It sets the working directory to be the directory containing the Deconvoluter.

Deconvoluter Version 1.00, 5/28/01
• Initial release version.