If I want to visualize the co-occupancy of a particular type of genomic data. Let's say DNA methylation/ histone modification with another set of data like TFBS sites e.t.c what's a good way to visualize such kind of data. Which types of plots or graphs would suit this analysis over a genome space.
One way to do this is to group your TFBS by transcription factor. For each factor, given the set of instances of that factor, calculate the average signal at each base across the factor (taking care to observe strand orientation of binding sites and reorienting signal, as needed).
If you don't want to reduce your data that much, you could generate a heatmap showing signal for each factor instance on its own row, using min-max or normalized min-max as a coloring scale.
You can permute rows randomly if you want to show a global pattern, or apply a sort order based on some statistic calculated over the instance's vector of signal. The sort order applied by a statistic may visually suggest an interesting biological phenomenon for a subset of instances of that factor, for example, particularly when associated with an indicator (on/off) heatmap of other data for the given factor instance (like genomic overlap with a ChIP-seq peak).
See the following figure for examples made from factors USF1, NRF1, P53 and BCL6, which use DNaseI cleavage and phyloP conservation as signals:
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
Hi Saad Khan, please provide some feedback on Manu's comment. Did the suggestion solve your problem?
IGV or make a density plot of abundance of reads over the site