What is the usage of CPM in edgeR analysis?

[setar...@gmail.com]

Hi Brian and all,

I'm confused about CPM in the edgeR analysis, I know that it's "count per million" and calculated as the raw counts divided by the library sizes and multiplied by one million. I was wondering if it is a kind of normalization? if your response is positive, so what about the TMM normalization and "calcNormFactors" function? what is the difference between these normalization types? 

All the best

[Brian Haas]

Hi Maryam,

CPM is similar to FPKM (or TPM) but doesn't take into account transcript length.  

TMM normalization is needed for cross-sample comparisons, regardless of whether your data are in CPM, FPKM, or TPM.

--
You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-users+unsub...@googlegroups.com.
To post to this group, send email to trinityrnaseq-users@googlegroups.com.
Visit this group at https://groups.google.com/group/trinityrnaseq-users.
For more options, visit https://groups.google.com/d/optout.

--

--
Brian J. Haas
The Broad Institute
http://broadinstitute.org/~bhaas

 

[setar...@gmail.com]

Thank you, Brian. Sorry, could you please let me know if the edgeR software convert counts to CPM and then analyze data? Similarly, I saw in several papers that the MA plot was drawn based on CPM value, x-axes is average logCPM, however, this axes is log counts in the MA plot generated by Trinity, please kindly tell me if these are equivalent to each other? is there any priority for using logCPM or log counts for MA plotting?

Thanks a lot in advance

[Tiago Hori]

I almost 100 percent sure edgeRnjas it's own internal cross library normalization algorithm, that's why you must provide raw count to it.

T.

Tiago Hori, Ph.D

Associate Director of Genomics 

The Center for Aquaculture Technologies 

Sent from my iPhone

--

You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-u...@googlegroups.com.
To post to this group, send email to trinityrn...@googlegroups.com.

[Brian Haas]

edgeR uses CPM, and the MA plots should be in log space, ideally average log(CPM) between the two conditions.  If that's not the case, I can look into it further.

best,

~b

On Wed, Aug 17, 2016 at 3:38 PM, Tiago Hori <tiag...@me.com> wrote:

I almost 100 percent sure edgeRnjas it's own internal cross library normalization algorithm, that's why you must provide raw count to it.

T.

Tiago Hori, Ph.D

Associate Director of Genomics 

The Center for Aquaculture Technologies 

Sent from my iPhone

On Aug 17, 2016, at 3:33 PM, setar...@gmail.com wrote:

Thank you, Brian. Sorry, could you please let me know if the edgeR software convert counts to CPM and then analyze data? Similarly, I saw in several papers that the MA plot was drawn based on CPM value, x-axes is average logCPM, however, this axes is log counts in the MA plot generated by Trinity, please kindly tell me if these are equivalent to each other? is there any priority for using logCPM or log counts for MA plotting?

Thanks a lot in advance

On Wednesday, August 17, 2016 at 10:05:06 PM UTC+4:30, setar...@gmail.com wrote:

Hi Brian and all,

I'm confused about CPM in the edgeR analysis, I know that it's "count per million" and calculated as the raw counts divided by the library sizes and multiplied by one million. I was wondering if it is a kind of normalization? if your response is positive, so what about the TMM normalization and "calcNormFactors" function? what is the difference between these normalization types? 

All the best

--
You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-users+unsub...@googlegroups.com.
To post to this group, send email to trinityrnaseq-users@googlegroups.com.

Visit this group at https://groups.google.com/group/trinityrnaseq-users.
For more options, visit https://groups.google.com/d/optout.

--

You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-users+unsub...@googlegroups.com.
To post to this group, send email to trinityrnaseq-users@googlegroups.com.

Visit this group at https://groups.google.com/group/trinityrnaseq-users.
For more options, visit https://groups.google.com/d/optout.

[Tiago Hori]

Right. But there is a per library normalization factor at some point, right? When you make the combined Matrix?

T.

Tiago Hori, Ph.D

Associate Director of Genomics 

The Center for Aquaculture Technologies 

Sent from my iPhone

To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-u...@googlegroups.com.
To post to this group, send email to trinityrn...@googlegroups.com.

[Brian Haas]

Yes, after running TMM you'll get the library size factor, and the cpm values are adjusted.

To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-users+unsubscribe...@googlegroups.com.

To post to this group, send email to trinityrnaseq-users@googlegroups.com.
Visit this group at https://groups.google.com/group/trinityrnaseq-users.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-users+unsubscribe...@googlegroups.com.

To post to this group, send email to trinityrnaseq-users@googlegroups.com.
Visit this group at https://groups.google.com/group/trinityrnaseq-users.
For more options, visit https://groups.google.com/d/optout.

--

--
Brian J. Haas
The Broad Institute
http://broadinstitute.org/~bhaas

 

--
You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-users+unsub...@googlegroups.com.
To post to this group, send email to trinityrnaseq-users@googlegroups.com.
Visit this group at https://groups.google.com/group/trinityrnaseq-users.
For more options, visit https://groups.google.com/d/optout.

[setar...@gmail.com]

Thanks a lot for your clarification.

On Wednesday, August 17, 2016 at 10:05:06 PM UTC+4:30, setar...@gmail.com wrote: