This is a test version of Biostars. For the public version, visit https://www.biostars.org.
RNA seq - TMM values

Hello!!

I did an RNA seq and then I did the normalization of TMM. Afterwards I chose:

  1. 1 contig which was over-expressed
  2. my 4 plants populations
  3. the TMM values regarding this specific contig and the 4 populations.

So I thought to do an ANOVA test to check for any st.sign. diff. among the populations for the specific contig. In the end I did a t test with Bonferroni adjustment.

Do you think that this whole thing I did make sense? If not what you could recommend to me?

Btw I tried the heatmap but I think my excel set up is somehow wrong...

Thank you very much in advance!

rna-seq tmm
• 3,360 views • link
updated by Ram Ram • written by Evlampia • Evlampia

Any reason why you did not use an algorithm such as DESeq2 or edgeR for your analysis?

• 0 views • link
written by WouterDeCoster WouterDeCoster

Then why the ANOVA test? I'm afraid I don't understand your question.

• 0 views • link
written by WouterDeCoster WouterDeCoster

Sorry. It s my fault. I am beginner in Bioinformatics. I did it with edgeR. And then I did an exact test analysis in Blast2GoPro based on some threshold for logFC, FDR, p value. And I took the contigs with FDR<0.05 as upregulated. Now I want to demostrate them as a figure. So now I can plot the TMM as a boxplot for my different populations and write on the top of my plot the FDR value. But I want also to show statistically significant differences btw the populations. The FDR is only about the contig. I dont have a unique FDR for each population. So my question is: How to show st.sign.differences btw the populations regarding the same contig? Sorry for the confusion....

• 0 views • link
written by Evlampia • Evlampia

I think the question was why didn't you go ahead and use edgeR/limma for the statistics part? Most people will use the linear modeling in limma/DESeq2 for statistical testing.

• 0 views • link
written by Matt Shirley Matt Shirley

I dont know what are these stuff...sorry...

• 0 views • link
written by Evlampia • Evlampia

Hi evlampia123x,

there are a few things that need clarification: What exactly do you mean by contig? Is contig = gene? So far, I understand that you did RNA-seq but which groups did you compare with edgeR? Do you have 4 populations under different conditions or compare 2vs2? What is the experimental setup? Please add some details to help understand your problem.

• 0 views • link
written by ATpoint ATpoint

I did RNA sequence using untreated plants coming from 4 different populations. I had 6 plants from each population. So overall 24 untreated plants. Then I took the results of the sequence and aligned them against a reference transcriptome. Then I did an BWA-MEM mapping. Then I took the raw data results and then I took the TMM (normalization) values (all these took place online in GeneData). When I say contig I mean contig which corresponds to a gene (let s say contig=gene). And I compared everything with edgeR (GeneData). Like this I took the TMM values. Are these infos enough? Sorry for the confusion. I am a beginner....if you cannot help me I will understand it...it s my fault bcs I cannot give you enough details because I dont know many things. I did the analysis with an other person...My question actually is: Is it correct to an ANOVA for TMM values or this is absolutely bulshit? Thank you guys for your help!!

• 0 views • link
written by Evlampia • Evlampia

0 answers

No answers yet.

Log in to answer this question.

More posts like this