Skip to content

Methylation – RRBS-seq and MeDIP-seq

December 31, 2011

A few months into my blog it seems like all I have been blogging about is ChIP-seq. I guess it has been what I have spent the majority of my time on since summer but I am performing quite a lot of other assays. Maybe I will blog about some of the protocols I have in the protocol section.

Anyway, to kick my ChIP-seq blogging addiction, I’ll move on the methylation and MeDIP-seq. I know MeDIP-seq is almost ChIP-seq but it a step in the right direction.

First, let’s talk about methylation. Unless you are looking at a really small genome or have free use of a HiSeq2000, there are essentially two assays to look at 5-meC methylation, i.e. reduced representation bisulfite sequencing (RRBS-seq) and methylated DNA immunoprecipitation sequencing (MeDIP-seq). Everything there is to be said about RRBS-seq on the wet lab side is said really well in this publication: The bioinformatics is another story. The software and workflows are still a little raw but I’ll save this discussion for another time. RRBS-seq is great because you get nucleotide resolution at the lion’s share of promoters and CpG islands and it only requires about 10 million reads so you can get 15 or so samples in one lane of the HiSeq. With MeDIP-seq you do not get nucleotide resolution but with the loss of resolution your theoretical coverage goes from 10% of the genome to 100%. MeDIP-seq requires about 50 million reads so sequencing costs are a little more. Neither method is superior and are completely complementary. I think there are some issues with the wet lab protocols that I have seen for MeDIP-seq but since the method is so similar to ChIP-seq the in silico workflows are further along.

One important thing to keep in mind is that the TruSeq adapters from Illumina are methylated, at least current batches are, so you can use them for RRBS-seq but not MeDIP-seq!!!


From → Posts

  1. Kalyan permalink

    Hi Ethan,
    I have been working on RRBS and had my first ever NGS library prepared using Illumina Truseq kit! I was looking for an estimate of the number of reads required for each sample. During the HiSeq2000 run, as the lanes are empty the technician has used my entire sample at various dilutions in each lane. We expected that we will get huge data. But surprisingly, due to high cluster density we have lost more than 60% of the data. Our local Illumina tech support personnel said that they had no idea of the recommended loading concentration on the lanes. I am curious to know if you had tried RRBS on HiSeq! Can u give us some inputs about the cluster density required in this regard!

  2. ethanomics permalink

    I am the wrong person to ask this question. I do the experiments, make the libraries and analyze the data but I send the samples out to be sequenced. That being said, library quantification is part art/part science. I recommend using qPCR. You can generate your own standards or buy them from Kapa Biosystems. You’ll probably save time and money if you buy them from Kapa as overpriced as they are. Then you can dial in how much to use, which will be a little different with different types of libraries. It is safer to underestimate then overestimate. But I think the key is to stick to a library quantification method so you know the optimal amount using the technique you use.

    The paper I referenced in the post says you only need 10 million reads per sample so you getting way more sequencing depth then you need. However, longer reads will get you more coverage.

  3. permalink

    Hi, Ethan,

    I am wondering if you have tried RRBS-seq with Truseq adapters. From the illumina website:

    TruSeq adapters are methylated. However, the kits are not currently compatible for the following reasons:
    WGBS: The PMM (polymerase mix) used in the TruSeq DNA Sample Prep kits cannot handle uracil bases in its template. This means that bisulfite-converted DNA cannot be processed directly with TruSeq kits, unless another polymerase is used to amplify the samples.
    RRBS: Same as above, and in addition the length of TruSeq adaptors results in the generation of adaptor dimers that overlap the size range of the targeted DNA, making them incompatible with this application.

    What do you think about it? Do you know if anyone has tried the Truseq adapters for RRBS?
    Thanks a lot!

    • ethanomics permalink

      Hi Josie,
      After traveling from Greece to California with a one-year old boy and missing a couple nights of sleep, I don’t have the energy to look at the RRBS-seq protocol now and I have never done RRBS or WGBS-seq myself. But from what I remember of the protocols, I would do a couple rounds or PCR before the agarose gel size selection step (for WGBS-seq you may just be able to do AMPure XP size selection if you are fragmenting with the Covaris). This converts the Y-shaped adapters to to dsDNA so it runs true to its size on an agarose gel. I address this issue in my ChIP-seq and MeDIP-seq protocols. I also highly recommend using my “Cycle Quantitation Protocol” to determine the correct number of cycles during the final PCR amplification. I think Kapa Biosystems is selling a polymerase for RRBS-seq and WGBS-seq amplification now (or they will sometime in the near future), which is worth looking into.

      On the number of rounds of pre-size selection PCR to do, it depends on how much DNA you have. Since I have so little DNA with ChIP-seq, I usually do 5 cycles. For MeDIP-seq, I have a lot more DNA so I usually just do 3 cycles.

      So to the question of what do I think about what Illumina has to say, it makes me think are these people stupid? How can the come up with such technology and be stumped such a trivial issue with library preparation? Did Bioo Scientific patent my pre-PCR idea? Illumina should have given me that job that I applied for (but thankfully they didn’t because it sounded boring).

      Hope that makes sense. Let me know if you need some clarification.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: