There are literally dozens of possibilities. R and Bioconductor are my first stop for microarray data; see limma for a good starting place. Commercial software such as GeneSpring, Partek, etc. are possibilities. GenePattern and BRB ArrayTools are a couple of open source solutions with GUIs.
If you haven't used any genomics software or R/Bioconductor, I'd strongly suggest you find someone local who has worked with microarray data. You say that you are starting with log2 values. It is likely that you will need to do some quality control, preprocessing and normalization, unsupervised analysis, and then supervised analysis, so having some local guidance is definitely the best way to get things done correctly.
That is usually called "finding differentially-expressed genes" and there are dozens of possibilities. Getting some local advice is warranted since that may affect which software is available and which you decide to use.
@ Sean: With all respect to your comment, "If a local help was available, why would I write a question on Biostar? ". No offences, but forums like seqanswers and biostars are meant to help the researchers who usually do not have any local help available. Furthermore, I was not seeking the definition of "finding differentially expressed genes", I guess that was quite clear from the heading of my question.
Point taken. As for "marking differentially-expressed genes", I am pretty sure that I misunderstood what you were asking so perhaps answered the wrong question. Feel free to clarify.
That is usually called "finding differentially-expressed genes" and there are dozens of possibilities. Getting some local advice is warranted since that may affect which software is available and which you decide to use.
@ Sean: With all respect to your comment, "If a local help was available, why would I write a question on Biostar? ". No offences, but forums like seqanswers and biostars are meant to help the researchers who usually do not have any local help available. Furthermore, I was not seeking the definition of "finding differentially expressed genes", I guess that was quite clear from the heading of my question.
Point taken. As for "marking differentially-expressed genes", I am pretty sure that I misunderstood what you were asking so perhaps answered the wrong question. Feel free to clarify.
Thank's Sean for your answer, data preprocessing is done. It is the step after that which is bit unclear