Dear Fatih,

In terms of number of genes: I assume all platforms should have nearly the same number of genes (or the same genes) on the slides.

Statistical design aspect: For miRNA microarray, each Agilent technologies’ cassette (slide) has 8 chips. If you have only case and control groups with 4 replicates, one slide would be enough and you would not have between slides variation but between chip variations. If you have 8 replicates for the same scenario, you can put 4 replicates for case and 4 replicates for control on each slide (two slides in total). If there is a serious handling difference between slides, you would be able detect it (this happened to us and I wanted to share this experience). As far as I know, other technologies have a single chip on each slide. I would assume this would increase between slide (chip) variations (due to handling) and would be hard to detect if there is a serious handling difference. I assume the above comments would be valid for microarray chips as well.

Software wise: Mostly array technologies have their own softwares but not free. Generally they would allow using for a limited time period (e.g. for a month). For miRNA microarray R has libraries for Affy and Agilent technologies. You had better check for the technology you are going to use before go for it.

Validating: In theory I agree with Christopher (RT-PCR). However, my understanding, there is different methods for confirmation (at least for miRNA microarray). You had better get advice from a molecular geneticist or ask from each platform what confirmation methods they have.

Have you considered ChIP-Seq by looking at various histone modifications that mark repressed, poised and active chromatin? Depends on your budget, if you're sending out samples or hoping for local analysis.

Consider to use 450K Illumina Infinium array to do it. We are working on it for several tissues in the past 3 years and we are obtaining nice results. We also compared the coverage of the CpGs to the Whole Genome Bisulfite Sequencing (WGBS), and it's worth to do it

You have to define the epigenetic changes you want to study first (i.e chromatin modifications, methylome, miRNAome...).

For chromatin modifications and methylome I suggest diagenode, they have optimized kits for ChIP and MeDIP, automated machines for a better reproducibility and less hands-on time and library preparation kits for very low input (50pg of DNA). They also provide ChIP-grade antibodies for the ChIP-seq. What is more their customer service is really good. You can also do BS-seq which provides single bp resolution.

For miRNAome illumina's kit for library preparation will allow you to make libraries from miRNA as well as the other non-coding RNAs in one experiment.

It would be a good idea to do some transcriptome analysis to prove the modifications you will find. For this I suggest illumina again.

Epigenome analysis using sequencing is expensive but, unlike microarray or qRT PCR, it allows you to have the complete landscape of epigenetic modifications.

Consider differential expression of miRNAs and genes (miRNA-seq and ssRNA-seq), I think it's a good starting point.

Illumina and CHIP-SEQ should provide a global view of the system you are interested in.

First of all you have to define which epigenetic changes you would like to to study. Then you could choose allumina or microaaray.

illumina is a good platform . We work with that

Illumina