How to filter constitutive expression genes and tissue-specific expression genes from RNA-seq data?

Question

[chesterlee7246]

As stated below,i have 4 tissue's RNA-seq data(FPKM),the following is my filter condition,is my way right?

---

constitutive <- subset(data,tissue1/tissue2 > 0.5 & tissue1/tissue2 < 2 & 
                               tissue1/tissue3 > 0.5 & tissue1/tissue3 < 2 & 
                               tissue1/tissue4 >0.5 & tissue1/tissue4 < 2 & 
                               tissue2/tissue3 >0.5 & tissue2/tissue3 < 2 & 
                               tissue2/tissue4 >0.5 & tissue2/tissue4 < 2 & 
                               tissue3/tissue4 >0.5 & tissue3/tissue4 < 2)

[Kevin Blighe]

No, it does not appear to be the right way (but it may be... it depends on what you are doing. Can you tell us more about the data that you have? For example, from where did you obtain it? What is the objective of your work?

[chesterlee7246]

I am a novice to bioinformatics.The data comes from 4 tissue of rice,it's fastq file.I aligned them to the reference genome then i got bam file.I obtained FPKM using stringtie.I want to filter constitutive expression genes and tissue-specific expression genes by FPKM,i am not sure whether it is feasible.Thank you!

[Kevin Blighe]

I see. So, you are not specifically interested in a differential expression analysis?

With your FPKM data from StringTie, my recommendation is to transform this data to Z-scale by using the zFPKM package. When your data has been transformed to Z-scale, you can take a look at the values in each sample:

• Z > 2 = expressed in sample
• Z < -2 = not expressed in sample

This way, you can construct lists of genes that represent each sample in a very simple way.