extract small heatmap from big one deeptools
2
2
Entering edit mode
7.2 years ago
biostarsb ▴ 30

Hello everyone,

I have a question. I had a heatmap with deeptools and I wanted to extract that small heatmap with the genes that have more signal (as shown in the image). Can someone help me?

Thank you in advance

http://pixady.com/image/jeo/

deeptools ChIP-Seq • 13k views
ADD COMMENT
0
Entering edit mode

waiting for help !!!!!

ADD REPLY
0
Entering edit mode

Please use ADD COMMENT/ADD REPLY when responding to existing posts to keep threads logically organized.

Are you losing money by the minute here :-) It has been less than 30 min since you posted the original question.

ADD REPLY
0
Entering edit mode

Tagging: Devon Ryan

ADD REPLY
2
Entering edit mode
7.2 years ago
Constantine ▴ 290

Run it again with the parameter --outFileSortedRegions myfile.bed . The genes you want should be in the first cluster that comes up (I assume you have run it with a --kmeans 4 ? )

ADD COMMENT
0
Entering edit mode

then i can extract the list of genes from the bed file and run again the computeMatrix then plotheatmap ?

Thank you

ADD REPLY
0
Entering edit mode

Sure ..should work this way

ADD REPLY
0
Entering edit mode

but the bed contain all genes not only those with high signal, for me i need to extract only those with high signal to do again onther computeMatrix and plotHeatmap

ADD REPLY
1
Entering edit mode

if you run again the plotHeatmap function with the parameters --kmeans 4 --outFileSortedRegions mynewfile.bed then your new bed file "mynewfile.bed" that is generated will be split into 4 different clusters. The genes you want are in the first cluster

ADD REPLY
1
Entering edit mode
7.2 years ago

You have a couple options. Firstly, you could save the sorted output from computeMatrix, gunzip it, take the top N lines, fix the header and then gzip it again (computeMatrix outputs a bit gzipped text file, where the first line is a header defining things like the matrix size and where samples are).

Perhaps a better options would be to use kmeans in plotHeatmap, pull out the BED regions for the preferred cluster and then run computeMatrix on that.

ADD COMMENT

Login before adding your answer.

Traffic: 2276 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