|
Steps of a microarray analysis
1. Probe preparation
To date, the Microarray Center uses long synthetic oligonucleotides or proteins as probes to generate microarrays.
The Center has no equipment for the high throughput preparation of PCR fragments from a cDNA clone library. However, it is possible to array any ready-to-use cDNA libraries supplied in standard 384-well titer plates.
2. Microarraying :
The Microarray Center currently uses commercial microscopy glass slides allowing covalent attachment of the probes. Several glass substrates are available depending on the application. Probe printing on membranes (macroarray) is also possible using the MicroGrid II microarrayer.
As an alternative to microarray preparation, several commercial oligonucleotide/DNA arrays are available .
3. RNA preparation and hybridisation
Several protocols and comercial kits exist for the extraction, the purification and the labeling of the target sequences with fluorescent dyes, and their subsequent hybridisation with the microarray. Depending on the amount of starting material available, an amplification step should be considered.
4. Microarray image acquisition and data analysis :
The readout of the microarray is captured as a colour coded 16-bit TIFF image using a dual-laser scanner for fluorescence detection. The scanner excites the fluorochromes associated with the target-probe DNA complex and measures fluorescence emission intensities at each DNA spot by scanning the microarray surface. The Genepix Pro software computes the intensities of the two fluorochromes at each location and provides a readout of the data as a spreadsheet linked to the image.
The next step in the analysis pipeline is the image gridding that allows to map the location of the pixels representing each spot. The image is then processed (normalisation, back-ground subtraction, signal to noise calculations etc) and the software calculates the intensity values of each probe as well as the quantitative ratio of the fluorescence intensity levels between the two samples being compared. These values provide an indication of the relative amounts of given RNA species in the samples that are representative of the gene expression levels.
5. Microarray data biostatistical analysis
Microarray experiments generate a tremendous amount of data that can be integrated and analysed using advanced algorithms employed by the Acuity 3.0 software. Here, the main analytical task is to reduce the complexity of the datasets so that trends and structure are revealed, and to group together genes or biological samples based on the similarity of their expression profiles. The fundamental assumption in the comparative analysis of transcriptomes is that coregulated genes often share common functions. By extension, possible roles for genes of unknown function can be suggested based on their association with genes of known function. Similarly, samples of unknown physiological status can be classified based on their association with samples of known physiological state.
There are numerous mathematical and statistical methods that can be brought to perform the analysis. The two main methods used in Acuity are Principal Components Analysis (PCA) and cluster analysis.
PCA is a data reduction technique. It reduces the complexity of a dataset by deriving a small number of variables from the data. The investigator then examines the behavior of genes/samples on a small number of these variables, instead of the behavior across many microarrays. The analysis thus provides a low-dimensional summary of the dataset easier to visualise graphically.
Cluster analysis is a grouping technique. It reduces complexity of datasets by partitioning data into a small number of sets. The investigator can then examine the behavior of each set, which is representative of the data in it, instead of the behavior of each genes or microarray. There are numerous algorithms to perform cluster analysis, some of them being used in Acuity (hierarchical and non-hierarchical methods, self-organising maps, k-means).
<<
| >>
|