Question

How to analyse RNA-based protein pulldown mass spectrometry data

0

Entering edit mode

barrypraveen • 0

@183232d3

Last seen 9 months ago

Czechia

Hi to all,

I did the RNA-based protein pulldown IP's for the mass spectrometry (without TMT labelling). I got the peptide counts and lot of them were zero. So i did log2(x+1) to all the data. since i have huge number of zeros (25%), it is influencing my data. My volcano plot from limma kind of flattened. There is a one major point is, i have zero peptide count in control but some values in expriment (e.g count is 5). I have to consider such values, because it is RNA based protein pulldown. So i did not expect huge number of peptide counts. I have IBAQ values and RAW peptide count as well. What will be the better way to analyse such data. can you please help me with this.

(I asked the same question in Biostars but i did not get any answer so i posted here)

Proteomics limma MassSpectrometry • 1.2k views

ADD COMMENT • link updated 12 months ago by Gordon Smyth 50k • written 2.8 years ago by barrypraveen • 0

score 0 · Answer 1 · 2021-06-26

0

Entering edit mode

Gordon Smyth 50k

@gordon-smyth

Last seen 8 hours ago

WEHI, Melbourne, Australia

I am not at all famliar with RNA-based protein pulldown mass spectrometry, so I can't tell you how to analyse it. My first suggestion would be to try analysing it like RNA-seq, e.g., with edgeR::voomLmFit or edgeR::glmQLFit.

ADD COMMENT • link 2.8 years ago Gordon Smyth 50k

0

Entering edit mode

First thanks for your suggestion. May i know the reason behind it. will it help in the normalization? Can you please explain me.

ADD REPLY • link 2.8 years ago barrypraveen • 0

0

Entering edit mode

I suggested RNA-seq pipelines simply because they are designed for counts and you have counts. The voomLmFit function is designed to be robust against lots of zeros. As I say, I am not familiar with the type of data you have so I cannot recommend anything with confidence.