BS-Seq from a single cell?

[eslondon]

Hello there,

is anybody attempting BS-Seq from a single cell or, anyway, from very limited amount of material?

Given that there is immunoprecipitation step like for Medip and MBD it should be doable?

Applying a similar logic to the one employed for single-cell transcriptomics, one would convert with bisulphite (quite a straight-forward chemical reaction which should not be impacted by low quantity of DNA), then purify (there might be issues there?) and use a WGA kit?

Any comments/ideas very welcome!

[simonandrews]

I've not seen any BS-Seq data from single cells (that's not to say no one is trying to do it though). Groups at our institute have done it with a few hundred cells, but going much lower than that would require some major changes to the current protocols.

There may be problems with BS-Seq which don't apply to (for example) single cell transcriptomics. Firstly you have to bisulphite treat before you do any amplification, since amplification will erase methylation marks. Also bisulphite treatment degrades your DNA so you'd need to be really careful to not just digest your whole sample away. Finally you'd need to be happy that your amplification was fair so that meth:unmeth ratios you saw in your library were the result of true methylation in the original samples and not biases in your transformation / amplification.

One of the reasons people are so excited about sequencers such as the PacBio or the (currently mythical) Oxford Nanopore, is that they can potentially get single molecule methylation data, something which is a huge technical challenge with current techniques.