[ECO]

Ariel,

Very nice. Do you have any objection to me featuring this in the bioinformatics forum as well?

Best,
Eric

[arielpaulson]

That's fine.

Thanks,
Ariel

[jyli]

Pingh-

Can you please advise where I can get Illumina DpnII tag library?

Thanks.

Jianying

[Tagz]

DpnII vs NlaIII

Hello!
I´m new to this topic and also have a question concerning the use of DpnII vs. NlaIII for solexa tag library generation. I´m interested in the gene expression analysis of mouse and aspergillus. Does anybody have an idea which enzyme fits better when analyzing both? I couldn´t find any information concerning the amount of NlaIII restriction sites in aspergillus.

What criteria are important for the enzyme selection? Amount of sites? ???

Thanks a lot.

[arielpaulson]

NlaIII vs DpnII in Mouse

Hi Tagz,

We've been digging deeper on this question lately, it turns out the difference between enzymes may not be so great as it appears. It is true that NlaIII generates twice the tag coverage of DpnII, at least in vertebrates. But in terms of how many genes are taggable, the difference isn't so huge, at least not in mouse. That, and DpnII site loss is greater in introns and intergenic space, where you don't need tags anyway.

We are also working in mouse and we cannot use Illumina's NlaIII kits because they are experiencing technical difficulties. I'm not sure if they're even selling them at the moment. But I know of people working with DpnII kits and apparently have no complaints, so we looked harder at the difference between enzymes. We used Ensembl 52, which has 31804 genes, of which 24133 can be uniquely or ambiguously identified by both enzymes (ambig indicates the tag occurs in a region of gene-gene overlap, NOT a repeat tag).

There are also 2929 genes which are uniquely taggable only by NlaIII and 1003 genes which are uniquely taggable only by DpnII. 3739 genes are not uniquely taggable by either enzyme.

There is an excel file which goes with that at ftp://ftp.stowers-institute.org/pub/...I_vs_DpnII.xls. You can inspect all 31804 genes and see which ones are taggable by which enzyme. If your top genes aren't DpnII-taggable, then you need a different protocol.

As for aspergillus, my first guess is it would look like S. cerevisiae. DpnII site prevalence is linked with CpG islands, so they shouldn't be under-represented compared to NlaIII sites in fungi, which don't use CpG regulation like vertebrates do. I don't have stats on that yet because the genome isn't in Ensembl format and I have not had the time to upgrade my scripts for months now But if Ensembl Fungi gets rolling before I do (and they have aspergillus on roll-out) then I can get exact figures.

Ariel

[Tagz]

Thanks Ariel.
This information is quite helpful. I guess it doesn´t matter that much which enzyme I pick.