Original version of this document can be downloaded at www.biologie.ens.fr/fr/genetiqu/puces/publications/arrayplot/arrayplot_tutorial.html

Arrayplot overview

Versions

How to load data files

A script is available to convert automaticaly genepix 3 output files to arrayplot files.         Public access at http://transcriptome.ens.fr/lgm_bioinfo/format_to_arp.php       This script is available from the download page and run in any computer with GNU software PHP4 installed.       Any other common file format can be added to the script on request.

Choose and name an experiment
 

If you click on "load file" button, a dialog box appears. Choose a file, click on the "open" button. Arrayplot asks you for the name of this dataset. This name will be displayed on left box (see multiple experiments display) and gene information (see view gene information). Default is the file name.

Multiple experiments display
 

By repeating previous step, you can load up to 10 experiments. You can hide an experiment by unselecting the checkbox associated to experiment name in left box.

Clear graph
 

Use "clear graph button" to discard experiments.

result analysis

Arrayplot displays a 2D plot where each point (x,y) is defined by x=cy3 intensity and y=cy5 intensity. Gene name is displayed by default and can be hidden using "label points" checkbox.

logarithmic axis
 

You can toggle between normal or logarithmic axis using "log axis" checkbox.

Display gene names in the graph
 

If the "label points" checkbox is selected, arrayplot shows for each point, the name of the corresponding gene (gene name if provided, otherwise gene id).

Zooming
 

You can zoom in a special area of the graph by selecting top left corner of the area and use left click while draging to the bottom right corner of the area.

Unzooming
 

To come back to a global view, repeat the same operation selecting bottom right corner first and draging to top left corner.

Interactive sliding view
 

You can slide the graph using right click + mouse move.

Changing visual significance limit
 

3 lines are present in the plot that represents by default ratio cy5/cy3 = 1:1, 2:1 and 1:2. If the line 1:1 is placed to the center of the cloud, points outside the 2:1 and 1:2 lines represent genes with significantly changed expression (2 fold induction or repression). This limit can be changed using the "induction" spin button. You can for example change this limit to 3 to draw lines where ratio=3:1 and 1:3.

Scale to 65k
 

By default, arrayplot changes axis minimum and maximum values to offer the best view of the cloud. If you check the "scale to 65k" checkbox, axis minimum will be set to 0 and axis maximum to 65600 which are the minimum and maximum values provided the microarray scanners (intensity level is coded with 16 bits, i.e. 65536 values). This option provides a synthetic view of the cloud and is helpful to access quality of the data set compare to others that have been previously examined using this standard.

View gene information
 

By clicking on a point of the graph, some information concerning the corresponding gene will appear in the text field situated at the bottom of the graph: dataset name, gene name, gene id, normalised ratio, and function information (as provided in the arrayplot file).

Go to external database page describing a gene

 Clicking on a point of the graph displays the corresponding URL (as provided in the arrayplot file) in the "go to" text field. Cliking on the "Go to:" button will launch a connexion to the specified URL via your default web browser.
Search for a gene

You can search for a point by entering a gene id or the gene name. A red arrow appears in the graph to show the corresponding point(s) position. If there is no corresponding point, an error message is displayed. This option is useful for housekeeping normalisation.

result normalisation

You can choose to normalise using spots supposed to be transcriptionally stable across the experiments. This can be housekeeping genes or genomic DNA spots. This is recommended when most genes in the dataset have altered transcription between samples. Gene set should be distinct from background and cover the dynamic range of scanner.     1) label unvariant genes using "search for a gene" option     2) use the "norm for a=" spin button to change the slope of central line and align it with the housekeeping genes. (preferably in log axis mode)     3) normalisation factor id is indicated under "norm for a=". Dividing cy5 intensity by this factor will lead to housekeeping genes ratio=1. Ratio will reflect real mRNA quantity difference.

How to normalise using mean of ratio?
 

If most of the genes present in your array are not transcriptionally altered between the two conditions (usually true in experiments using whole genome arrays), you can normalise according to the mean or median of whole genes.     1) label unvariant genes using "search for a gene" option     2) If most of your genes are transcriptionnaly invariant, you should obtain, before normalisation, a cloud forming a line. Use the "norm for a=1" spin button to change the slope of central line and align it with the cloud. (preferably in log axis mode)     3) normalisation factor id is indicated under "norm for a=". Dividing cy5 intensity by this factor will lead to housekeeping genes ratio=1. Ratio will reflect real mRNA quantity difference.

Why do I have a cloud that I can't align?
     - you observe a x or y shift of the whole cloud: your array present a high background noise only in cy3 or cy5.
     - you observe 2 distinct linear clouds: your array presents local high background noise only in cy3 or cy5.
     - you observe an asymptotic limit in high intensities: your laser power was too high and measure is saturated. With common microarray scanners, intensities > 60000 should be discarded.

Printing graph

Arrayplot offers a convenient printing interface. Using this tool, you will be able to adjust printing quality, orientation, margin size...

Saving graph image

You can choose to save your image as a bitmap image (BMP) or as a vectorial image (WMF). While BMP only saves a pixel representation of the graph, WMF save the coordinate of each lines and points of the graph, allowing zooming without any loss of quality.

Other tools in Arrayplot

purpose	The pie display allows to get a rapid synthetic view of gene intensities and ratio. Ratio is displayed as red (cy5) and green (cy3) radius. Size of the pies correspond to the sum of cy3 and cy5 intensities. This tool was constructed to analyse quickly 96 measurements in one screen. The tool is limited to 150 measurements because for higher population, pies are too small and not easy to interpret.
how to change pie size or geometry?	Using the "pie size" button, you can change pie size in order to see low intensity spots. The "pies/line" button allows to change the number of pies per line. Using the "pie size" button, you can change pie size in order to see low intensity spots. The "pies/line" button allows to change the number of pies per line.
how to normalise using housekeeping genes?	If you have some housekeeping genes, you can use the "norm/cy5" button to adjust their ratio to one (half of the pie green, half red).

Scanarray info viewer
 

The "ScanArray infos" of tool menu permits to retrieve informations encoded in TIFF images header (chanel scanned, laser power, barcode...)

Calculator
 

A simple link to windows calculator (shortcut keys: 'control'+'k' )

Installing / uninstalling Arrayplot

Install program is available for download at www.biologie.ens.fr/lgmgml/publication/arrayplot/download.php The installation is fully assisted, just copy the executable to you PC and run the program.

Arrayplot ini file
A file called arrayplot.ini will be created in your windows file to save Arrayplot options. Options can be edited using the "options" menu.
This file contains the following sections:
[fichiers]
    format=fichiers_Arrayplot
    default_directory=H:\uller.pc\résultats puces\
[graph]
    label=on
    animated_zoom=true
[label_font]
    charset=0
    height=-13
    name=Times New Roman
    size=8
[pies]
    nb_printing=false
    nb_printing=true

How to give a feed back?

Arrayplot Homepage

How to cite Arrayplot?

Papers referencing Arrayplot:
Ammonia pulses and metabolic oscillations guide yeast colony development, Devaux et al., submitted
New insights into the PDR network through the characterization of YRR1 transcription factor regulation system , Le Crom et al., submitted
A repair-replication checkpoint activated by ionising radiation in S. cerevisiae, Mercier et al., Mutation research, dec 2001 [medline]
An artificial transcription activator mimics the genome-wide properties of the yeast PDR1 transcription factor , Devaux et al., EMBO reports, june 2001 [medline]
Book: DNAmicroarrays in neurobiology, Potier et al., CRC Press, 2001
Book: Developement of microarrays to analyze gene expression in brain extracts and single cells, Potier et al., MIT Press, 2001

Philippe MARC- LGM - ENS- Paris last updated: 17 december 2001