The normalised Deseq2 data show hits for my knockout gene in the data generated by the total RNA sample from KO mutant.

Question

[sri]

I am very new to this world of RNA seq data analysis. I performed RNA seq between wildtype and Gene of interest (GOI) Knockout (KO) mutant to determine the effect of GOI in differential gene regulation. I am using Galaxy for the data processing. My experimental pipeline is (a trial and error one) >> Trimmomatic >>Hisat2 >>Htseq count >>DeseQ2 analysis(No intermediate steps in between) . I have one big question that worries me a lot at the moment. In the normalised count data, I could see value 12 for the GOI in GOI KO RNA sample whereas other two replicates show 0 in its respective place for the gene. What could be the chances of it??. It bothers me a lot. please help me.

[ATpoint]

Are the three samples based on the same clone or are these three independent KOs? How was the KO achieved (CRISPR?). Complete deletion of the gene or a partial deletion of some exons?

[sri]

Yes, all the replicates belong to the same clone. The knock out is performed by simple replacement of the GOI by hygromycin resistance cassette using homologous recombination method. I am working with Ustilago maydis, a basidiomycete fungus which has high homologous recombination rate.

Answer 1

[ATpoint]

Ok I see. So the "value 12", is this the output of counts(dds, normalized=T), so not log2? If so, it should be fine as this is a very low count and probably not to worry about. You should compare it with counts for genes that you know are not expressed in this cell type and some that are highly activate to get a feeling. A bit of background noise is not unusual and does not necessarily argue against a successful KO. Does the KO gene come out as significantly lower in the KO condition?

[sri]

The log2FC value for the KO gene is -1.22907. In the wildtype situation, the count values are 1300,1116, 2000 in respective replicates but in KO condition as I told the values are 12,0,0. I am sorry if it doesn't answer your question

[ATpoint]

It's all fine. The 12 is just noise I would say and nothing to worry about. To increase your confidence, just check some genes that are canonically not expressed in this type of fungus. You'll also see non-zero but smallest counts.

[sri]

Thank you very much for your help:):)