good avgprof file, but heatmap is blank with plot with -R exon or -R bed

[xiaor...@gmail.com]

Hi ngsplot author,

    Very appreciate you wrote this amazing program to plot large scale sequencing data.

    But my problem is that when I plotted with your example files at exon region or bed region, there was heatmap file but blank and avgprof file was fine. When I plotted at tss region, both avgprof and heatmap were fine. So, what's going on here?

     Command line: ngs.plot.r -G hg18 -R exon -C hesc.H3K27me3.1M.bam -O H3274me3atexon -D refseq -FL 300

                            ngs.plot.r -G hg18 -R bed - E AAA.bed -C hesc.H3K27me3.1M.bam -O H3274me3atexon -D refseq -FL 300

     Whatever I used peak.summit.bed to be AAA.bed or custom exon.bed, the results were the same.


    Thanks!

Rui

[xiaor...@gmail.com]

When I replotted with the .zip file, the result was the same.

Rui

[xiaor...@gmail.com]

I also try -G hg19 -R exon with example file, still no heatmap file, but -R genebody works fine.

Rui

On Sunday, March 2, 2014 2:44:29 PM UTC-8, xiaor...@gmail.com wrote:

[ningyi shao]

Xiaorui,

Our example bam is a very small bam file for testing the installation of ngs.plot, and only a very small part of the genome was covered. So if you want a more comprehensive heatmap coverage, you need to feed a normal bam file to ngs.plot.

Bests,

Ning-Yi

--
You received this message because you are subscribed to the Google Groups "ngsplot-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ngsplot-discu...@googlegroups.com.
To post to this group, send email to ngsplot...@googlegroups.com.
Visit this group at http://groups.google.com/group/ngsplot-discuss.

To view this discussion on the web visit https://groups.google.com/d/msgid/ngsplot-discuss/4f78e38e-3ecc-4404-8d7e-3314b29d768e%40googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

[Li Shen]

It's very likely that only a very small number of regions contain actual data. The average plots are dominated by these regions. H3K27me3 is a histone mark that does not seem to enrich at exons. All these factors matter. 

I don't think it is a bug of the program. Rather, it is data related.

Li

--

You received this message because you are subscribed to the Google Groups "ngsplot-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ngsplot-discu...@googlegroups.com.
To post to this group, send email to ngsplot...@googlegroups.com.
Visit this group at http://groups.google.com/group/ngsplot-discuss.

To view this discussion on the web visit https://groups.google.com/d/msgid/ngsplot-discuss/d4a12cb1-65e2-4a7a-9a5f-948d01d37d2d%40googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

--

Li Shen, Ph.D. Computer Science
Assistant Professor
Department of Neuroscience
Icahn School of Medicine at Mount Sinai
New York, NY 10029
#################
ngs.plot - the coolest NGS plotting tool: https://code.google.com/p/ngsplot/

Want ChIP-seq differential analysis? Use: https://code.google.com/p/diffreps/

[xiaor...@gmail.com]

Hi Dr. Li and Dr. Shao,

     Thank you for your quick response!

     Actually, I first tried ngsplot with one of my real dataset with -R bed (peak.summits.bed which contains more than 80,000 peaks and the bam file is huge, >2G), and the other 2 .bam files (64M and ~400M) with -R exon,  but the result were the same. So I started to try example file to check where the problem was.

     Thanks again!

Rui

On Sunday, March 2, 2014 2:44:29 PM UTC-8, xiaor...@gmail.com wrote:

[xiaor...@gmail.com]

Sorry, Dr. Shen and Dr. Shao.

Rui

On Sunday, March 2, 2014 2:44:29 PM UTC-8, xiaor...@gmail.com wrote:

[xiaor...@gmail.com]

Actually, I can see the value for each column in hm1 file, not zero. It seems there is some problem to generate heatmaps. And I will try to use the value to generate heatmaps with treeview. If it is possible I can send you the zip file for you to take a look.

Another question is : Is it possible to generate heatmaps for [chip subtract input] which is well accepted for final singal for chip-seq? I can do it manually with your ngsplot and treeview, but it is too trivial.

Thanks!

Rui

On Sunday, March 2, 2014 2:44:29 PM UTC-8, xiaor...@gmail.com wrote:

[ningyi shao]

Hi XiaoRui,

Would you mind to provide more details of your issue?

Such as the details of your datasets, the reference genome you used, the header of your bam (samtools view -H), and one zip file.

You may email the zip file to us by private email, and we will keep it in confidence.

Bests,

Ning-Yi

--

You received this message because you are subscribed to the Google Groups "ngsplot-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ngsplot-discu...@googlegroups.com.
To post to this group, send email to ngsplot...@googlegroups.com.
Visit this group at http://groups.google.com/group/ngsplot-discuss.

To view this discussion on the web visit https://groups.google.com/d/msgid/ngsplot-discuss/18121e40-c771-465a-aebd-7ce22630dea6%40googlegroups.com.

[xiaor...@gmail.com]

Hi Dr. Shao and Dr. Shen,

    I just try 10000 peak summits to plot the same bam file, and it works very well. Maybe >80000 peak summits are too much to draw the heatmap and the height is not enough to draw so many points. Is there any threshold I can tune? Will replot.r -RR 1 work on this huge dataset? Any way, I will try this. This problem may happen to -R exon as well, because exon number is huge too.

    And I will send you the zip file later on. Actually, my adviser Xiang-Dong Fu in UCSD knows Dr. Wei Wang very well.

Rui

[xiaor...@gmail.com]

Hi Dr. Shao and Dr. Shen,

    I just try replot.r -RR 1, 2, 8, still not working.

[Li Shen]

Hi Rui,

The -RR means the larger the value, the more compressed the heatmap. So I think you want to increase that value instead. 

The color scale matters a lot to a heatmap. If the automated scale does not work for you, you can try to set a custom value. 

Li

--

You received this message because you are subscribed to the Google Groups "ngsplot-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ngsplot-discu...@googlegroups.com.
To post to this group, send email to ngsplot...@googlegroups.com.
Visit this group at http://groups.google.com/group/ngsplot-discuss.

To view this discussion on the web visit https://groups.google.com/d/msgid/ngsplot-discuss/1b9398e4-c305-4f93-b048-b08062c316d6%40googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

[xiaor...@gmail.com]

Hi Li,

     I just try replot.r -RR 100   >  not working

                               -RR  200 > not working

                               -CO green > not working

                               -CO Blue > not working

                               ngs.plot -G hg18 -R genebody   > working

          head or taill -n 10000 peak.summits.bed > 10000peak.summits.bed

                               

                           then try:

                               ngs.plot -G hg18 -R bed -E 10000peak.summits.bed    >  working

                               ngs.plot -G hg18 -R bed -E 40000peak.summits.bed   > working

                               ngs.plot -G hg18 -R bed -E 60000peak.summits.bed   > working

                               ngs.plot -G hg18 -R bed -E 65000peak.summits.bed   > working

                               ngs.plot -G hg18 -R bed -E 70000peak.summits.bed   > not working
                               ngs.plot -G hg18 -R bed -E 80000peak.summits.bed   > not working

                               ngs.plot -G hg18 -R bed -E 20000tailpeak.summits.bed  > working

                               ngs.plot -G hg18 -R bed -E 30000tailpeak.summits.bed   > working

    So it seems that there is limitaion for the bed file for -R bed, can't be more than 70000; is it possible to tune this? because the number of confident peaks for the target protein is ~88000. Thanks!

Rui

[ningyi shao]

Hi Rui,

Would you mind to paste all console output when you try ngs.plot -G hg18 -R bed -E 70000peak.summits.bed?

And would you mind to send us an example zip file of ngs.plot?

Ning-Yi

To view this discussion on the web visit https://groups.google.com/d/msgid/ngsplot-discuss/a56da988-91db-476e-a5dd-c3f0505fecbe%40googlegroups.com.

[xiaor...@gmail.com]

Hi Ning-Yi,

    I just sent you the output files. By the way, ngsplot is ultrafast and versatile. Thanks!

Rui

[xiaor...@gmail.com]

Hi Ning-Yi,

    Finally, I figure it out! The PDF software problem. When I use Acrobat, it is fine but not working with pre-installed document viewer.

    Thank you so much! No problem right now! I love ngsplot! The results are pretty good!

Rui

[ningyi shao]

Hi Rui,

I am happy everything is OK. You are welcome!

Ning-Yi

To view this discussion on the web visit https://groups.google.com/d/msgid/ngsplot-discuss/54b3de18-c7e0-4fd8-9bf2-28db71754a19%40googlegroups.com.