wyguo / ThreeDRNAseq

A pipeline for differential expression and differential alternative splicing analysis
https://github.com/wyguo/ThreeDRNAseq
GNU General Public License v3.0
60 stars 27 forks source link

[Help] Duplicated biological replicate for Controls #33

Closed RxLoutre closed 2 years ago

RxLoutre commented 2 years ago

Hello, sorry it is me again.

With others smaller experimental design I have tried, I can use 3DRNAseq with no issues. But with this one particular dataset, I get an error that I do not understand. Here is what the experimental design look like :

ID,Condition,Time,Accession,Replicate,Quant_file
1,Priming,12,Col,1,1_Priming_12_Col-0_1
2,Priming,12,Col,2,2_Priming_12_Col-0_2
3,Control,12,Col,1,3_Control_12_Col-0_1
4,Control,12,Col,2,4_Control_12_Col-0_2
5,Priming,12,Pla,1,5_Priming_12_Pla-0_1
6,Priming,12,Pla,2,6_Priming_12_Pla-0_2
7,Control,12,Pla,1,7_Control_12_Pla-0_1
8,Control,12,Pla,2,8_Control_12_Pla-0_2
9,Priming,48,Col,1,9_Priming_48_Col-0_1
10,Priming,48,Col,2,10_Priming_48_Col-0_2
... 
#etc until sample number 40

Replicates here are biological replicates.

When I load this file in 3DRNAseq, if I don't use the "technical replicate" button, it does not want to load and show the following error :

Duplicated biological replicate labels: Control.X12.Col, Control.X12.Pla, Control.X24.Col, ... Please double check (1) the replicate labels in your metadata table or (2) whether sequencing or technical replcates were selected.

I have doubled check many times my experimental design file and the lines the error are pointing toward, but I don't see the issue. Is there something I have missed ?

Thank you for your help,

Best,

RxLoutre commented 2 years ago

Oh no, this have been bogging my head for 2 days but I think I understand the issue. The experiment itself was kind of split in 2 and had two controls with similar conditions, but they are for different experiments.

ID,Condition,Time,Accession,Replicate,Quant_file ... 3,Control,12,Col,1,3_Control_12_Col-0_1 4,Control,12,Col,2,4_Control_12_Col-0_2 7,Control,12,Pla,1,7_Control_12_Pla-0_1 8,Control,12,Pla,2,8_Control_12_Pla-0_2 11,Control,48,Col,1,11_Control_48_Col-0_1 12,Control,48,Col,2,12_Control_48_Col-0_2 13,Priming,48,Pla,1,13_Priming_48_Pla-0_1 14,Priming,48,Pla,2,14_Priming_48_Pla-0_2 15,Control,48,Pla,1,15_Control_48_Pla-0_1 16,Control,48,Pla,2,16_Control_48_Pla-0_2 17,Control,12,Col,1,17_Control_12_Col-0_1 18,Control,12,Col,2,18_Control_12_Col-0_2 19,Control,12,Pla,1,19_Control_12_Pla-0_1 20,Control,12,Pla,2,20_Control_12_Pla-0_2 29,Control,24,Col,1,29_Control_24_Col-0_1 30,Control,24,Col,2,30_Control_24_Col-0_2 31,Control,24,Pla,1,31_Control_24_Pla-0_1 32,Control,24,Pla,2,32_Control_24_Pla-0_2

As you can see here are all the control. And 3DRNAseq saw several times the condition Control.12.Col.

In my opinion, I should analyze the two experiment separately, because the controls have the same conditions, but were sampled at different times. Do you also agree or would you suggest something different ?

Best,