The GenQuery Engine (GQE) is a utility that takes columns of Affymetrix probe sets (or any list of GenBank or TIGR Unique IDs) and:

1) Automatically retrieves gene-specific information from internal (Affymetrix) and external (TIGR, UniGene) databases.

2) Generates gene-specific hyperlinks to multiple databases.

3) Outputs an LUT-based score for probe sets called "Increased" or "Decreased" in GeneChip-based comparisons of AffyMetrix Mu11K chips.

Most of the features can be independently disabled (e.g., if the user wants the program to run faster, web-based (TIGR/UniGene) queries can be disabled).

The latest version of the GQE has been updated to recognize the new Entrez-based UniGene format, and the links to TIGR databases (ET*, TC*, HT*) have all been repaired.

The GQE is featured in the following articles:
Mills J.C. and Gordon J.I. (2001) Nucleic Acids Res. 15, e72 [Full Text PubMed]
Mills J.C., Roth K.A., Cagan R.L., and Gordon J.I. (2001) Nat. Cell Biol. 3, 175-178 [Full Text]

The GQE is a Microsoft Excel ".xla" or "Add-In" file. Once it is downloaded and uncompressed (from "zip" or "StuffIt" format), you can open it on any computer running a version of Microsoft Excel. Because the GQE is a macro, you will be warned that the GQE file contains macros. You will have to press <Enable Macros> to run GQE. The rest of program operation is designed to be user-friendly and includes help about the various features.

Requirements: The GQE was written in Visual Basic for Applications on PCs running Windows NT and Windows98 in Excel2000. It has not been tested (but should work) in Excel97 for Windows.

Macintosh users: The Mac version of GQE is no longer supported. Please let us know if there are any of you still out there, and we can discuss options.

Experienced Add-In Users: The GQE works best when either opened like a workbook from within Excel or when opened directly (i.e. when the user double-clicks on the GQE icon to start GQE and Excel). If the user manually installs the GQE as an Add-In from within Excel, it will work but will not add and remove itself from the "Tools" menu, and it will launch every time Excel is opened. On the Mac, the GQE does not work at all when installed manually as an Add-In.

PC Users:
Click here to download GenQuery Engine and LUTs for the PC.
(Last updated November 16, 2003)

Macintosh Users:
Click here to download GenQuery Engine and LUTs for the Macintosh.
(Last updated on February 14, 2003-- NOTE: NO LONGER SUPPORTED)

NOTES on the GQE:

The GQE will take any list of rat, human, or mouse GenBank (e.g."AA12345") or TIGR IDs (e.g. "TC12345" or "ET12435") in a Microsoft Excel file and retrieve gene-specific information from TIGR and/or UniGene. This is useful for automatically updating gene lists with the latest UniGene information. Gene lists do not have to have come from Affymetrix experiments. However, to get the GQE to recognize them, users must add an underscore "_" to the end of each UID (this can be done using Excel's " Concatenate" worksheet function).

The GQE was updated in September, 2001 to enable better functionality with "U" style chips and the corresponding Affymetrix internal databases. The program now recognizes "U" style Affymetrix IDs (e.g. "987654_f_at") and will query a separate Excel worksheet (designated as an "Internal Database" and prepared by the user from the list of probe sets and descriptions supplied by Affymetrix) to return the Affymetrix description for each probe set (Descriptions for "U" style chips all begin with "Cluster Incl"). It will then extract and insert a hyperlink to GenBank and query UniGene with the GenBank ID. The program will also recognize a column containing Affymetrix Descriptions of the "Cluster Incl" type and will extract the GenBank ID for hyperlink preparation and UniGene querying.

The LUT scoring module of the GQE has only been tested on Mu11K and Mu6500 chips, but it should work, in theory, with "U" style chips as well. LUT scores generated this way will still have relative value (i.e., a score of "6" will indicate more reproducible data than a score of "1"), but they will not correspond in any absolute sense to the analysis presented in the Nucleic Acids Research paper cited above. Also, users should vary the scaling factor they input into the program until a reasonable distribution of scores is achieved (e.g., around 40% of probe sets scoring >3). We suggest doing this, because the LUT system is based entirely on the raw Average Difference and Average Difference Change signal intensity, and "U" style chips will behave differently in this respect from Mu11Ks. We are considering preparing a revised U74 series of LUTS.


Outline of program and notes on how it evolved

GenQuery Engine (GQE), ©2001-3, by Jason C. Mills and Jeffrey I. Gordon

Please send questions or comments to:

Dept. Molecular Biology/Pharmacology
Washington University School of Medicine
660 S. Euclid Ave.
St. Louis, MO 63110