General Array Procedure

From Experiment to data.

RNA Extraction

Purification of Total RNA
There are several methods of RNA purification. Hot phenol extraction is an economical, and easy method for getting total RNA.
Purification of mRNA
Most labelling strategies utilize mRNA. There are several methods of purifying mRNA from Total RNA. Price Comparison
oligo-dT spin columns   (PDF)
Pharmacia sells oligo-dT resin pre-packed into columns for rapid spin purification. Each column uses up to 1.25 mg of Total RNA and yields around 10-25 ug of mRNA depending on the experiment. These kits used to be economical, rivaling the price of oligo-dT resin, until Amersham raised the price. 4 columns now cost around $160.

NEB oligo-dT aliquots   (PDF)
NEB sells a slightly more economical method of mRNA purification. The kit contains oligo-dT resin pre-hydrated and aliquoted into eppendorf tubes for batch binding. Spin filters and buffers are included for elution. 8 purifications costs about $150.

oligo-dT Resin
The standard method for mRNA purification. It's easy and affordable. A vacuum manifold, or spin filters can greatly decrease the purification time.

Qiagen oligotex beads   (PDF)
The kit is easy to use but is somewhat costly.

Labelling cDNA

Reverse Transcription with amino-allyl dUTP
The aa-dUTP protocol is the easiest, most economical, most reliable, and best method for labelling cDNA made from either total or mRNA. It involves reverse transcription in the presence of aa-dUTP which can later be conjugated to either Cy3 or Cy5. It offers high specific activity and more even probe labelling than incorporating the labelled nucletides into cDNA directly.
RT with Cy-dNTP
Reverse Transcription with teh direct incorporation of Cy labelled nucelotides is widely used by most labs, but becoming displaced by the method above.

3D Dendrimers with Cy dyes
A company called Genisphere makes DNA spheres of crosslinked nucleotides with Cy molecules attached. Each sphere can hold approximately 250 flours. The spheres come with a sequence that can hybridize to a tag sequence incorporated into the cDNA. The method is sensitive such that small amounts of total RNA can be used. However it is expensive and finicky.

Slide Hybridization and Washing

3X SSC, High temp.
The standard method for hybridizing microarray slides utilizes 3-4X SSC, and temperatures of 63-65 deg Celsius.
Formamide Based Hybs
Similar to High temp hybs except you can add formamide and lower the hyb temp accordingly. Some people prefer pre-blocking their slides with BSA in the hybridization buffer.

Scanning and Quantification

Axon Scanner and GenePix
The Axon scanner is fast becoming one of the most popular scanners. It scans both channels at once so it's very rapid. GenePix software is a breeze to use.

GSI Scannaray 3000
This scanner from GSI Luminomics was one of the first to hit the market, and it is quite ubiquitous. It scans arrays one channel at a time. The output is a Tiff image for each channel. It usually comes bundled with a software package called Imagene for quantifying the scans. This scanner has a problem producing scans that overlay perfectly (they're usually a few pixels off in one direction). The overlay problem is taken care of by Axon GenePix. However in Scanalyze one can see the image misalignment, when it ocurrs for a particular scan (the problem is somewhat intermittent).

Scanalyze is Michael Eisen's freeware program for quantifying array scans. It's available at It runs on windows computers.

The ArrayWorx Scanner is a high resolution CCD based system for scanning arrays. Any of a number of wavelengths are available, so you're not stuck to Cy3 and Cy5. The data analysis takes place using an SGI workstation.

Other Software, Number Crunching, and Beyond

Joe DeRisi and cohorts at the annual Cold Spring Harbor Micrarray course wrote this simple, Perl based flat file database. It's called AMAD (Another Micro Array Database). The great thing about AMAD is that it is simple, it works, and it is extensible.

Coming soon to, the DeRisi lab will be releasing NOMAD, a bigger, better, more efficient implementation of AMAD using MySQL and Java.

Serafini Lab/Oracle Database

Number Crunching
Bibliography of Array Statistics papers

The AMAD Project

AMAD is a simple way to access your array data from any computer. SInce it is written in Perl, it is very easy to develop Modules for doing whatever analysis you can think of. Once you develop something, it can be used by others (if you choose), and it's likely that someone will think of a new idea with your code, improve it, and submit it back to the AMAD site for you to download and use. And so the software can grow via the collective IQ and become useful and available to everyone.

Chris Seidel