I was told to obtain a chip seq peak figure from one of the articles (http://www.cell.com/cancer-cell/abstract/S1535-6108(12)00489-8). Figure 1B. My following questions might be dumb but even though I know some stuff about this workflows, this terminology usually creates a glitch on my mind. Before asking questions, I would like to explain that what I did so far.
- I downloaded chip seq data (BED files)
- Ran macs14 peak calling workflow. There are control files of the each unique run. (Like
GSM698575_7_ChIP_Input.bed) Therefore instead of omitting the control in macs14, I entered this control bed file too.
I got macs output(below) and optional ( bed graph or wig file). I ran it two times to see the difference btw bedgraph( I transform to bigwig with kentutils) and wig files. ***If I go with the bed graphs, I have to use kent utils to clip my beds then transform to bigwig with bedgraphtoBigWig.
macs14_MACS_bedGraph macs14_diag.xls macs14_model.pdf macs14_model.r macs14_negative_peaks.xls macs14_peaks.bed macs14_peaks.xls macs14_summits.bed
I couldn't get a similar graph in the figure 2. ( I know that this paper called the peaks with SWEMBL). So,
My Questions are:
- Shouldn't I get similar peaks as figure 2
- What additional step I should do?
- My cannot observe same graphs with bed graph and wigs. Why can I get the same graph ? Which file type should be my input of IGV?
- Was using control wrong?
Any advice will be priceless for me. Thank you for your help and patience for reading this post.