Go Back   SEQanswers > Bioinformatics > Bioinformatics

Similar Threads
Thread Thread Starter Forum Replies Last Post
cufflinks reports extremely high FPKMs for short transcripts cram Bioinformatics 10 11-24-2014 04:24 PM
Cufflinks "randomly" dropping fpkms -- how to replace? arielpaulson Bioinformatics 5 02-21-2013 08:58 PM
Cufflinks reporting differnt FPKMS for the same gene nwalwort Bioinformatics 6 02-09-2012 07:47 AM
cufflinks 1.0.3 missing FPKMs mgogol Bioinformatics 3 07-29-2011 11:08 AM
Cufflinks low FPKMs and other wonders yehudithasin RNA Sequencing 0 07-04-2011 05:23 AM

Thread Tools
Old 02-22-2011, 02:47 PM   #1
Location: Seattle, WA

Join Date: Sep 2010
Posts: 17
Default Normalization of FPKMs from Cufflinks

Hi there,
I have been using the Tuxedo suite tools , specifically I have been using bowtie, tophat,and cufflinks. Now that I have FPKM values I have a couple of questions.
1. What are the most common ways folks are using to normalize their FPKM data?
2. I am looking into differential expression of the FPKM values across my samples and time points. I have successfully run cuff compare but when I run cuffdiff I get results (Tested id XLOC_000001, locus, etc.) but my gene ids are not in the output. any suggestions people have on why that might be missing? I can always map the tested id with cuff compare results. Any suggestions people have are apprciated. Thanks
greener is offline   Reply With Quote
Old 02-23-2011, 04:28 AM   #2
Location: Israel

Join Date: Oct 2010
Posts: 19
Default missing gene id


Regarding question #2 -
If in the annotation gtf file you provide for cuffcompare there is no "gene_name" field, then in cuffdiff output there are no gene names.
For example, when I used ucsc gtf file it has no "gene_name" but Ensembl does have this field.

Regarding question #1 -
As far as I understand, FPKM values are already normalized. but I am not sure about this.

reut is offline   Reply With Quote
Old 02-23-2011, 06:45 AM   #3
Location: Seattle, WA

Join Date: Sep 2010
Posts: 17

Thanks Reut, It was my understanding that one's individual cufflinks run generate normalized FPKMs. I have run multiple cufflinks runs across conditions, time points, and with replicates. I'm curious how folks normalize sets of cufflinks results. Additionally aside from cuffdiff have folks had much luck using LIMMA on FPKMs for differential expression? Thanks!
greener is offline   Reply With Quote
Old 05-06-2014, 05:55 AM   #4
Senior Member

Join Date: Jul 2013
Posts: 142

anybody got answer?
My trouble is to make csHeatmap of normalized diff gene's fpkm values.
For example, csHeatmap can be generated using XLOC IDs (cummeRbund gets the real fpkm values and makes heatmap in Log scale). However, to get scale +1 to -1, I have to mannualy get all fpkm values and then convert to Log10 and then normalize them and, finally, generate heatmap using gplots.
Is there any other method which can do it automatically as csHeatmap(cummeRbund) does with Log scale Or any way to normalize using cummeRbund. How others are doing ?
Some solutions Please
thank you
jp. is offline   Reply With Quote

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 05:52 PM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2020, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO