msbentsen
commented
1 year ago Hi @xflicsu
We have not validated TOBIAS on CUT&Tag data, so I cannot say whether it works or not. In theory, it should work, but there might be a few potential issues:
- If the CUT&Tag is for H3K27ac, any Tn5 footprints will be exclusively found in H3K27ac-rich regions
- In the same context, you might have to provide peaks/"regions of interest" that are larger than the peaks you would classically find for CUT&Tag, as you are also looking slightly outside of the signal region.
- I don't know how the Tn5 bias is affected by the CUT&Tag protocol, as ATAC-seq uses Tn5, but CUT&Tag uses pA-Tn5
So I would say, if there is already a tool developed specifically for CUT&Tag, it's better to use that. Otherwise you might try out TOBIAS and see if the CTCF shows a visible footprint like shown in the paper above. That might be a good test to see if it worked. Hope that is helpful!
xflicsu
Hello TOBIAS developer, I read the paper published by Steven Henikoff. They show "CUT&Tag protects a “footprint” spanning 80 bp directly over the CTCF motif". And the similar footprints were obtained using different salt concentration washes in their Fig. 5c.
Furthermore, Guo-Cheng Yuan developed the CUT&RUNTools to perform CUT&RUN processing and footprint analysis.
Both ATAC-seq and CUT&Tag are based on Tn5 enzyme digested enrichment experiment. I wonder whether the TOBIAS is suitable for the CUT&Tag of H3K27ac for footprinting analysis?