Hello, I keep getting empty results from differetial expression analysis tool.
I have input like this
Describe Samples.tabular
expression matrix like this
Please help. Thank you!
Welcome, @Yuqiao_Min
Thanks for posting all the details! Super helpful.
It seems that you need to adjust the inputs so that the IDs inside the files “match up”. That is how the tool knows how to combined the data into a new output.
Notice how all of the sample/condition names are all OneWord in the example. Tools prefer that data keys are in that format – most have trouble with spaces.
Also notice how certain columns should be exactly the same format between files. That is what I mean by a “match” between files. Tools also prefer exact matches 
.
All of these may the same thing to a person but entirely different things to a computer: Condition A, condition A, conditionA, Conditiona, conDitionA.
Scroll down on the tool form to the help section for more details about what any particular tool expects. This is the help for this tool:
Help
Trinity assembles transcript sequences from Illumina RNA-Seq data. This tool performs differential expression analyses on a transcriptome assembled with Trinity.
Inputs
This tool uses the matrix produced by ‘Build expression matrix for a de novo assembly of RNA-Seq data by Trinity’ tool.
You must describe your samples and replicates with a tabular file looking like this:
| ConditionA | CondA_replicate1 |
|---|---|
| ConditionA | CondA_replicate2 |
| ConditionB | CondB_replicate1 |
| ConditionB | CondB_replicate2 |
| ConditionC | CondC_replicate1 |
| ConditionC | CondC_replicate2 |
| ConditionC | CondC_replicate3 |
This file can be generated with the ‘Describe samples and replicates’ tool. It will probably be the same file as used in the tool ‘RNASeq samples quality check for transcript quantification’. The names in column 2 must match the names given in the tool ‘Build expression matrix for a de novo assembly of RNA-Seq data by Trinity’.
You might need to back up and re-do upstream steps. Or, you can adjust the files you currently have as long as they are plain text. Galaxy hosts a bunch of utilities for text manipulation. Most are named the same as command-line tools if you are familiar with those … or you can review our “Cheat Sheet” here: https://training.galaxyproject.org/training-material/topics/introduction/tutorials/data-manipulation-olympics/tutorial.html
Please give that a try, I think it will solve the current issue. For this specific tool – removing the space in Column 1 of the sample file is probably enough.
@jennaj Hi, do you mean deleting the space in column 1 in describe sample file?
I tried to do that but I still got empty results. Is there anything else that I should change? Thank you1
Hi @Yuqiao_Min
Yes, that new file format looks fine.
Why I think that is “fine”
- Two columns of data
- No spaces
- Unique values in each column
- The second column in this file contains the same values as column header values in your prior expression matrix
And … no results could mean no expression differences, or differences that couldn’t be calculated at all. You only have two samples, and DESeq2 usually requires at least two samples in two conditions to calculate certain scientific values. So, a minimum of four samples. That is just how the tool works – anywhere – not just in Galaxy and not just in this specific tool. Not having enough samples may not be trapped well by the tool. Maybe read up more about the usage? Links are down in the help section.
So – try to figure this out then I can help more with the specifics once you know that the inputs are minimally appropriate for what this tool is expecting to process.
And, there are other ways to do transcriptome DE!
I would suggest starting here, and looking at one of the Intro tutorials near the top. One is for when you have a reference genome available and one is for de-novo prediction. These don’t involve a true assembly, and instead a predicted coordinate/footprint method. https://training.galaxyproject.org/training-material/topics/transcriptomics/