Hi guys! I'm experiencing troubles about microarray gene expression data clustering. I have lot of breast tumor data to cluster together. Since there are a lot of clustering techniques for microarray gene expression data I really don't understand what is the best algorithm able to fit my data. I just need a good quality and reliable review or paper able to compare all the available methods with explicitly mentioned advantages and disadvantages about using one technique instead of another. Can anyone help me please?
Clustering is a classification method that is applied to data, it predates bioinformatics by a good deal and the choice of clustering really depends on the data and its properties as well as the hypotheses that need to be tested.
There are probably no review articles specifically on clustering in the way that would be helpful to you. This is not a novel scientific methodology that would discussed in such papers. There are however a very large number of textbooks and training materials on clustering and their applications. For example a google search on clustering microarray data turns up the following:
and many others. As you search you may find many overly technical discussions, feel free to ignore those and find something easier to digest.
One aspect that is not always emphasized in tutorials is to remember that clusters are an approximation and an idealistic and simplified representation of the reality. Therefore multiple types of clustering results could produce contradictory statements - yet all those could all be valid within the bounds of the approximation.
Here is a quick link for clustering tutorials in R and shows how the output looks like for the similar kind of data. You can choose the best fitting clustering method based on the properties of your data which Istvan has alread clarified a bit here!!