pllotting tag density across TSS and TES

Question

[kanwarjag]

I am plotting Chipseoq to be more precise Medip seq data across TSS when I use ref seq TSS file to plot the tage density across TSS (1000bp either direction) and TES I found the plot is showing TSS and TES very close almost 500 bp. I feel something is wrong as in the published plots e.g Fig 2D of http://www.sciencedirect.com/science/article/pii/S0092867415005607

Is there a rule of thumb to decide how much distance we should be selecting across TSS.

Thanks

Answer 1

[Bioradical]

Try using deepTools to plot this sort of data. You would be using a BED file of your gene data which can be obtained from UCSC, and a bigWig file of your TF or target of interest. Using the computeMatrix command go ahead and use the scale-regions mode and then follow the manual for information on parameters. The use the heatmapper command to plot this data (metagene plot) as well as develop a heatmap.

[kanwarjag]

Thanks Carlos,

The problem is if you see in Fig 2D of http://www.sciencedirect.com/science/article/pii/S0092867415005607

both TSS and TES are plotted that is what I want to plot but deep tools would not do same. I used TSS and gene body file to compute matrix, it is not giving me TES but just give TSS. Am I doing something crazy here?

[Bioradical]

You are probably using reference-point mode, scale-region mode is what you're looking for. It'll give you the TSS and TSE assuming your files have this data.

https://github.com/fidelram/deepTools/wiki/All-command-line-options

This link should give you all the command line options to generate the plot.

Also, ngsplot is another useful tool I use when doing this type of analysis. This requires a bam file and a bedfile of regions (or you can choose some preincluded regions such as exons, tss, genebody from ensemble or refseq annotation).