volcano plot of up and down genes into two parts, resprctively.

Question

[Bing Dwen Dwen]

When I do volcano plot of my RNA-seq data, I found the up and down genes cluster into two parts (two parts in up genes and two parts in down genes), resprctively. Can someone please explain to me why this is the case? Thanks a lot.

[cpad0112]

strange that entitites with fc values between ~5-~10 fold changes, on either side, are missing or not statistically significant.

[Bing Dwen Dwen]

Do you know why it happens？ some mistakes during processing of data？

Answer 1

[ATpoint]

These have large fold changes but rather low pvalues. If I had to guess then these are genes with many zeros. Some prefiltering of the data, e.g. edgeR::filterByExpr() can help removing unreliable calls.

[Bing Dwen Dwen]

You are right :) there are some zore data inside, so it means that are these data with zeros unavailable?

[ATpoint]

What do you mean with "unavailable"?

[Bing Dwen Dwen]

Do you mean that there are a lot of zero data in the data, will these data be removed after filtering the data?