4 replies to this topic

[nisot]

I am wondering if there is a very quick and easy way to estimate if there is a difference in the methylation status of gentically identical DNA from different tissues. I am thinking of a restriction enzyme digest with methylation sensitive and non MS isoschizomers and comparing the banding patterns, however we are talking about the whole mouse genome so although my supervisor thinks its a great and simple idea, I am not so sure as I would expect incoherent smears on the gel for both samples, plus we would need tons of DNA and replicates.
Does anyone have a better idea that is quite simple and doesnt involve any clonning 2D blots or any other cumbersome methods?

I am thinking of doing the digest followed by PCR (then what?), will this work? if yes could someone please elaborate or give links for this protocol. All suggestions welcome. THanks guys you will really be helping keep my sanity (my supervisor is ridiculously optimistic )

[methylnick]

hey Nisot,

you would probably find out that there are many ways to skin this cat and it all goes back to what you want to ask, what methylation status are you trying to look at? how precise do you need it and do you really need to know the locations of the methylation changes or not?

This would then determine whether you do an HPLC, HPCE approach which just gives you the methylation level of the genome as a whole, or you can get more refined to do a genome-wide sequencing screen to look at not only the changes across the whole genome but then you can start to answer where in the genome the changes have occurred.

With restriction enzymes, although a great way to test methylation changes, you are limited to where the recognition site sits in the genome and how many there are, that is why you see combinations of enzymes to increase the coverage.

hope this helps somewhat in making your decision.

Nick

[nisot]

methylnick, on Dec 28 2009, 09:23 PM, said:

hey Nisot,

you would probably find out that there are many ways to skin this cat and it all goes back to what you want to ask, what methylation status are you trying to look at? how precise do you need it and do you really need to know the locations of the methylation changes or not?

This would then determine whether you do an HPLC, HPCE approach which just gives you the methylation level of the genome as a whole, or you can get more refined to do a genome-wide sequencing screen to look at not only the changes across the whole genome but then you can start to answer where in the genome the changes have occurred.

With restriction enzymes, although a great way to test methylation changes, you are limited to where the recognition site sits in the genome and how many there are, that is why you see combinations of enzymes to increase the coverage.

hope this helps somewhat in making your decision.

Nick

Thanks for a swift response Nick! we are planning on doing a proper whole methylome study as a follow up to this imaginary quick and dirty method. Basically, we need to determine if other types of methylation occurs in our DNA and if so, incorporate this into the more precise whole methylome method. so I guess HPLC type methods wouldn't work. I was thinking of using REs that recognize Cs in any contexts and comparing the difference in bands generated. I dont really care what the seq of the bands, just wanna see if there is a difference....

Edited by nisot, 28 December 2009 - 04:01 PM.

[MoB]

It seems to be not that quick, it seems to be definately not dirty. Do you mean something like that?


Differential Methylation Hybridization

Best

MoB

[nisot]

MoB, on Dec 30 2009, 09:01 AM, said:

It seems to be not that quick, it seems to be definately not dirty. Do you mean something like that?


Differential Methylation Hybridization

Best

MoB

That method like most others is CpG biased! my supervisor has finally realised that there is no QnD method and we will proceed straight to the main method. BATMAN for MeDIP is being updated to remove its CpG bias