SEQanswers

Go Back   SEQanswers > Bioinformatics > Bioinformatics



Similar Threads
Thread Thread Starter Forum Replies Last Post
Finding the Similarity/dissimilarity of RNAseq samples using correlation dena.dinesh Bioinformatics 2 06-09-2015 12:42 PM
how to visualize a few hundreds RNAseq samples in IGV/GenomeBrowser shirley0818 RNA Sequencing 0 01-27-2015 07:28 PM
legend for heatmaps ashuchawla Bioinformatics 4 04-09-2014 04:30 AM
Wierd overlapping paired ends with bimodal paired end distance distance distribution Mouth_Breather Illumina/Solexa 12 08-24-2012 12:06 PM
CanFam2 data RNAseq samples and GenomicRanges biofreak Bioinformatics 0 08-08-2011 11:35 AM

Reply
 
Thread Tools
Old 07-08-2015, 05:59 AM   #1
chudar
Junior Member
 
Location: Europe

Join Date: Jul 2015
Posts: 9
Default Different dendograms for correaltion heatamp and distance heatmaps for RNAseq samples

Hi,

I am new to analyze the RNAseq data. I have 6 samples (3 Controls and 3 treated) for which I would like to find how similar/dissimilar are they.
I found the similarity/dissimilarity between samples using Eucledian distance as well as Correlation matrix but the resulting dendogram are different between this two methods. I cant understand why this is happening. Kindly guide me.

Code for heatmap based on Euclidean Distance

Code:
vsd <- varianceStabilizingTransformation(dds, blind=TRUE)
heatmap(as.matrix(dist(t(assay(vsd)))))
Code for Correlation heatmap
Code:
heatmap(cor(assay(vsd)))
chudar is offline   Reply With Quote
Old 07-08-2015, 06:11 AM   #2
dpryan
Devon Ryan
 
Location: Freiburg, Germany

Join Date: Jul 2011
Posts: 3,480
Default

Given that the correlation and euclidean distance are different, I wonder why you expect them to produce identical dendrograms.
dpryan is offline   Reply With Quote
Old 07-08-2015, 06:19 AM   #3
chudar
Junior Member
 
Location: Europe

Join Date: Jul 2015
Posts: 9
Default

Thanks for your reply. So in order to find which samples are close to each other, which is better to use. Correlation matrix or distance matrix? Kindly guide me
chudar is offline   Reply With Quote
Old 07-08-2015, 06:28 AM   #4
dpryan
Devon Ryan
 
Location: Freiburg, Germany

Join Date: Jul 2011
Posts: 3,480
Default

Neither is better than the other. You simply need to choose which one you prefer. I generally use euclidean distance, but one of my colleagues uses cor(). Just make sure to indicate which, though you tend to get similar results either way.
dpryan is offline   Reply With Quote
Old 07-08-2015, 06:39 AM   #5
chudar
Junior Member
 
Location: Europe

Join Date: Jul 2015
Posts: 9
Default

Thanks for your guidance. . Now I understood the logic.
Is my approach for using cor() on assay(vsd) with following code is right?

Code:
vsd=varianceStabilizingTransformation(dds, blind=TRUE)
Code:
heatmap(cor(assay(vsd)))
chudar is offline   Reply With Quote
Old 07-08-2015, 06:43 AM   #6
dpryan
Devon Ryan
 
Location: Freiburg, Germany

Join Date: Jul 2011
Posts: 3,480
Default

It looks OK by eye-ball, but I should warn you that I rarely use the heatmap() function manually, so I might easily miss something obvious there.
dpryan is offline   Reply With Quote
Reply

Tags
correlation, distance matrix, euclidean

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off




All times are GMT -8. The time now is 01:50 PM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2019, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO