This depends on the library and the use of the library. If a chip is being used to assess general expression changes in a mammalian species, for example, an appropriate positive control might be a housekeeping gene from the (specific) mammalian genome under discussion and an appropriate negative control would be non-mammalian (e.g. Arabidopsis). However, if the chip were to be used to assess inter-species variation within the same genus, for example, controls should be chosen to assess similarities and differences between those species. In general, there are controls which most libraries would have in common.
The most common controls include:
(a) a positive control (usually a spike) to control for labeling, hyb, chip quality etc;
(b) a hyb control (can be the same as (a) or separate) to assess variation as a function of position;
(c) a negative buffer control to assess intensity in the absence of DNA (and as a control for wash efficiency/carryover);
(d) a negative, non-printing, blank control to provide a convenient way to assess non-local background properties (and used to fill in space in the grid);
(e) landmark controls to control for grid placement and annotation alignment. back to top

What are the advantages to using single spotting over duplicate spotting on a microarray, and vice-versa?

With single spotting you can get a higher density, a better usage of the library (ie a greater number of slides per batch per inking) and better use of time.
The theoretical advantage of duplicate/replicate spotting is that you should get more confidence in the intensity measurements. However, you usually can't afford to print more than 2 spots per slide (because of aforementioned limitations in density and spots per visit). This means that you don't have much choice for statistics - no standard deviation, no means of choosing one value over another if there is a large discrepancy etc.
If you are in a position where you can go with more than 2 spots, you are faced with where to distribute them on the slide in an unbiased way. Adjacent spotting doesn't take into account hybridization variation and so you might as well just print bigger spots (or scan at higher resolution)!
In practice, we have found that because of these limitations and probably more importantly because the increase in data quality does not warrant it, it is much more sensible to go with single spotting. The majority of variation arises from hybridization and labeling which swamps within-slide replicates variability. Far better to increase data quality confidence by technical (and biological) replicates ie. slide replicates and assess positional variation using hybridization controls. back to top

How many replicates are needed for a microarray experiment?

This depends on what is to be achieved with the experiment. If the experiment is to give a first impression of some scenario, a few replicates or even none might be a good starting point. To be able to use statistical methods, e.g. Student's t-test or ANOVA, one needs at least two replicates. In practice three or more give more precise predictions. The more replicates one can generate, the better the prediction is going to be.

There are two types of replicates, technical and biological ones. In technical replicates the same sample is hybridized on several slides. In biological replicates several samples with the same treatment are hybridized to several slides. Technical replicates are generally the exception for cases where biological replicates are very hard to come by. back to top

How much total RNA is required for labeling ?

The total RNA/DNA required for labeling is recommended as:

in-house printed arrays: 5-10 ug total RNA, 100-2000ng if amplification is required

Agilent expression arrays: 50 ng - 5 µg total RNA

Agilent CGH arrays: 0.5 to 3ug gDNA

We have DNA ready to print. Can you print it for us?

Yes. Please contact us to ensure your samples are optimized in order to guarantee the best possible microarray. back to top

Where can I find information about the constitution of the arrays?

Click here for downloadable files. back to top

How can I find out if my gene of interest is on your microarray?

Click here for downloadable gene files. back to top

Can I add clones to your array?

Yes, please contact us for instructions on how to proceed with the addition of a clone(s) to a library. back to top

Are the microarrays ready to use?

Yes, the arrays can be picked up or distributed by the facility. back to top

Do I need to block the array before use?

No. It is not necessary. back to top

Which labeling protocols do you recommend?

Please see our Labelling, Hybridization, Scanning page. back to top

Should we label with each sample with both fluors? What is the difference between dye-flip and Universal RNA?

Yes. Using a dye-flip design allows you to deal with labeling bias (ie. unequal binding affinities due to the added tag). Universal reference RNA fulfils this requirement when labeled with one dye and hybridized to the same microarray as each experimental RNA sample labeled with the other dye. In this way, the signal for each gene can be normalized to its own unique factor allowing comparisons of gene expression across multiple samples. back to top

How soon must I use the microarrays once they've been printed?

Although it's better to use the microarrays as soon as possible after printing, we are confident in using microarrays up to 6 - 8 months old. We recommend keeping the arrays in a dry and dark enclosed space. back to top

On what substrate are your microarrays printed?

This depends on the type of array. We use the substrate that is the most appropriate for your application. back to top

How are the array quality control checks carried out?

The array manufacture quality process is described below.
• All slides are inspected manually to identify any major printing issues.
• One slide every 20 to 30 slides is hybridized with labeled random 9-mer oligonucleotide and scanned using ScanArray express scanner.
• Presence/absence, shape, signal intensity vs. background, diameter and DNA binding site capability are measured for each spot on the slide using files generated by Imagene software.
• A batch report is generated by an algorithm that combines all information collected for all slides' quality control. The batch report is sent to users. Position and description of 'flagged' spots (spots absent or thought to be unusable during post hybridization analysis) are listed on a separate file, also provided to users.
• Dust, sub-grid defects and other noticed irregularities are recorded.
• As an option, we offer to hybridize your RNA using our standard detection method
• We can also perform QC hybridizations using a sample of your RNA or possibly RNA from our libraries. back to top

Do you have scanners that I can use?

Currently the facility has four scanners in operation. An Axon Instruments GenePix 4200AL, a Perkin Elmer ScanArray Express, a Applied Precision ArrayWoRx, and an Agilent DNA microarray scanner. These can be booked for use through a sign-up sheet (you may contact us and have a time booked in advance). back to top

When scanning, should I set the PMT settings the same for both images?

No, settings should be set to acquire the best possible image for each independently. Normalization of data during the analysis stage will handle differences in channel intensity. back to top

Do you have a template file for gridding with GenePix?

Yes, grid files for use with GenePix can be supplied upon request. back to top

I have noticed that spot size can vary from sub-array to sub-array as well as within each sub-array. Does this mean there has been a problem with the spotting?

No. Spot size can vary from sub-array to sub-array because each sub-array is produced exclusively by one particular pin used on the 48-pin arraying tool. Each pin has its own spotting "behaviour" which, among other things, results in a characteristic average spot-size for each sub-array. For the quill pins we use, this size is nominally between 130-160 microns. back to top

In addition to this, a given quill pin can often generate smaller or larger than average spots on a given array or batch of arrays. This is because of a whole series of other factors such as variations in salt content of the sample (due to subtle evaporation differences across the plates while they are exposed to the environment) and variations in the DNA size and concentration in sample wells. back to top

However, spot size is not an issue from a QC'ing point of view unless the size of the spot falls below a threshold for which it is not possible to obtain sufficient numbers of pixels to accurately measure the spot's intensity. Again, with respect to spot shape, the same rules apply - the shape is irrelevant as long as there are sufficient pixels to quantitate. There may be issues with some imaging software being unable to accurately quantitate spots of unequal size or of strange shape, but this is something that most software developers have integrated into some of the more sophisticated imaging systems. back to top

What if I have intensity signals where the 3 X SSC buffer spots are?

This happens from time to time randomly for unknown reasons. Stringent testing has shown that this does not infer sample contamination or printing carry-over. back to top

What sort of data will the facility provide?

The data provided will depend on the users needs. At the minimum, a full printing batch quality control report will be provided along with a gene ID text file which includes the physical location of the genes on the array, basic printing information, and associated annotation. Laser scanning generates .tiff images which can then be quantified via software such as Imagene. The output of the quantification is a spot intensity report in a text format which can also include information on background, basic analysis and flagged areas of interest. Data analysis results created by software such as Genespring, Bioconductor, etc. can be saved in various formats depending on the users needs (i.e. import into pathway analysis software, visualization software, etc.). back to top

Can I obtain help with the analysis of my data?

Yes, we offer full training and consultation on the analysis of your data. This includes software suite instruction as well as statistical methodology direction.

Can I publish my data to a public repository such as GEO ?

Yes, many investigators are publishing their data to the public repositories such as ArrayExpress at the EBI and GEO at NCBI which are designed to accept, hold and distribute MIAME compliant microarray data. A similar database CIBEX is under development in Japan. The array information (platform) is uploaded by us if it is not already published, then the user can upload their sample data along with the specifics of the experiment. Our data analysts can help you with this process and make sure your data is MIAME compliant and acceptable to these repositories. back to top

What normalization method do you recommend ?

Generally it is preferable to use a LOWESS normalization. The applicability of the LOWESS method depends on having enough probes on the slide, around 1000, and the slide not having a strong bias to certain gene groups. If e.g. a specialized cancer chip is hybridized with tumor sample in one channel and healthy tissue in the other, the genes measured with the slide are globally turned on in one of the channels and a global normalization method like LOWESS would fail. back to top

How do I acknowledge the Microarray Facility when I publish my results?

Please see our Prostate Centre Microarray Facility Acknowledgment Policy page for details

What happens if I get a defective chip purchased from your facility?

If the default can be linked to a printing event from our facility, we'll replace it back to top.

What is the cost of various services offered at the Microarray Facilty?

Please contact our manager: Stephane Le Bihan. back to top