Integration of Cut&Tag to ATACseq and RNAseq data - impact of a mutation
1
0
Entering edit mode
24 days ago

Hello,

I am studying a mutation in the gene Spi1 versus the wildtype. We generated ATACseq and RNAseq, and saw a clear correlation between both (x = FC ATAC, y = FC RNA expression). In this post I represented the comparison between Homozygote and wildtype for my mutation.

enter image description here

I ran seacr to call peaks in Cut&Tag in stringent mode for different histone marks (H3K27ac; H3K27me3, H3K4me1) and my gene of interest Spi1. I would like to cross those data with ATACseq that we generated previously.

One idea would be to look at the regions differently open in ATACseq between mouse of different genotypes, and then look in those regions the normalized level of reads in each genotype. Does this idea make sense ? Or should I reason more in terms of differential binding for each marks and then cross those genomic regions ? What would be a good idea to integrate all those data ? Any examples or link to article ?

I have included on the graphe the differential binding at the peak of ATACseq position of my gene of interest SPI1 but I'm unsure for the next steps... Any help is really appreciated...

enter image description here

RNAseq Cut_Tag ATACseq • 255 views
ADD COMMENT
2
Entering edit mode
24 days ago
LChart 4.8k

It seems to me you're thinking mechanistically about this PU.1 mutation. In general it is hard to disentangle direct effects (PU.1 -> Accessibility -> Expression) from indirect effects (PU.1 -> Expression of chromatin remodeler -> Expression), except that where PU.1 does not bind you can only have indirect effects.

For display of these kinds of data, I have recently been using alluvial plots to display differential status of genes or gene sets:

enter image description here

and grouping peaks (or gene promoters when RNA is included) by trajectories (e.g., "+.+-" for up, nc, up, dn). This is effectively a "side view" of peak annotation. One thing you could do is take consensus peaks (or simply bins) and cluster them (or run ChromHMM) which will identify potential groups, and you can assess whether there are differential patterns within each group as a secondary analysis.

ADD COMMENT

Login before adding your answer.

Traffic: 4861 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6