# Normalization -TMM

**URL:** <https://forum.depmap.org/t/normalization-tmm/4122>\
**Category:** Q&A\
**Tags:** omics\
**Created:** [April 6, 2025, 6:21pm UTC](https://forum.depmap.org/t/normalization-tmm/4122 "2025-04-06T18:21:46Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![AMBREEN](https://avatars.discourse-cdn.com/v4/letter/a/aeb1de/32.png) [@AMBREEN](https://forum.depmap.org/u/AMBREEN)\
**Post date:** [April 6, 2025, 6:21pm UTC](https://forum.depmap.org/t/normalization-tmm/4122/1 "2025-04-06T18:21:46Z")

</div>

Hi  
I want to understand that if I have to do TMM normalization rather than using TPM as TPM is not considered a very good parameter to compare across different samples. Which file with Raw reads would be appropriate to carry out theTMM normalization?

---

<div class="post-metadata">

**Author:** ![simz](https://yyz1.discourse-cdn.com/flex035/user_avatar/forum.depmap.org/simz/32/440_2.png) [@simz](https://forum.depmap.org/u/simz)\
**Post date:** [April 11, 2025, 8:46pm UTC](https://forum.depmap.org/t/normalization-tmm/4122/2 "2025-04-11T20:46:37Z")

</div>

Hi,

I’d suggest using OmicsExpressionGenesExpectedCountProfile or OmicsExpressionTranscriptsExpectedCountProfile which contains expected counts estimated by RSEM instead of raw reads.

As a heads up, in the upcoming release we will update our RNAseq pipeline which will introduce changes in expression data. Please stay tuned for our communications and files descriptions for more details.

Best,  
Simone

---

<div class="post-metadata">

**Author:** ![AMBREEN](https://avatars.discourse-cdn.com/v4/letter/a/aeb1de/32.png) [@AMBREEN](https://forum.depmap.org/u/AMBREEN)\
**Post date:** [April 16, 2025, 8:47am UTC](https://forum.depmap.org/t/normalization-tmm/4122/3 "2025-04-16T08:47:50Z")

</div>

Thanks Simone. This is helpful.  
Please help me with another query as well – The data present in OmicsExpressionGenesExpectedCountProfile or OmicsExpressionTranscriptsExpectedCountProfile are log 2 transformed or general reads ?

---

<div class="post-metadata">

**Author:** ![simz](https://yyz1.discourse-cdn.com/flex035/user_avatar/forum.depmap.org/simz/32/440_2.png) [@simz](https://forum.depmap.org/u/simz)\
**Post date:** [April 16, 2025, 1:17pm UTC](https://forum.depmap.org/t/normalization-tmm/4122/4 "2025-04-16T13:17:30Z")

</div>

They are not log2 transformed.