RNA-Seq Data analysis

Question

[singlecell_bio]

Hello everyone, I had tried comparing two gene expression datasets in the .tsv file formats (reads quantified using Salmon) that were obtained from the RefineBio website. I was able to annotate the data using Galaxy using a GTF file from ENSEMBL, but when I tried using edgeR or limma, an error would occur. I was wondering if there is another way to compare datasets and build a volcano plot to show the differentially expressed genes between two samples.

RefineBio link to data: https://www.refine.bio/experiments/SRP117268/transcriptome-analysis-reveals-determinant-stages-controlling-human-embryonic-stem-cell-commitment-to-neuronal-cells

The expression files for each day (day0-22) are available. My goal is to compare day 0 with days 2, 4, and 6. I want to make volcano plots for each comparison and find the differentially expressed genes.

Thank you in advance !!

[seidel]

You can't use edgeR directly on the data quantified with Salmon because it generates abundance estimates (usually real values), whereas edgeR requires integer read counts (not the same thing). However, I see there's a catchSalmon function in edgeR now for reading Salmon data...did you try that?

[singlecell_bio]

Thank you! I'll look into that. I also have a .csv file format of the same experiment with FPKM data. Can the FPKM data be used in any way to find differentially expressed genes between the two datasets?

[dsull]

Unfortunately no. Your problem is that there are no replicates and there's really nothing you can do that will allow you to do differential gene expression analysis and obtain p-values.

The most you can do is some exploratory analysis: e.g. you can make a heatmap of some genes you're interested in across the different days, you can look at the FPKM values and see which genes are more highly expressed than others within a given sample, etc. But you can't do a differential gene expression analysis between two conditions (i.e. you won't be able to claim "these are the genes that are statistically significantly differentially expressed between day 2 and day 0").

[singlecell_bio]

Thank you for your insight!

Answer 1

[dsull]

You can't make volcano plots nor can you perform a statistical differential gene expression analysis, because there are no replicates! There is only one day 0 sample, only one day 2 sample, etc.

[singlecell_bio]

But is there a way to look at the differences in gene expression between the two days? I also have a .csv file format of the same experiment with FPKM data. Can the FPKM data be used in any way?