FlashLFQ normalization issue

[shrivastavabhishek]

Hi @trishorts @lonelu

I was running the FlashLFQ using cmd line and was expecting the normalised peptide intensity value in the QuantifiedPeptides.tsv file. I am sharing the parameters setting, PSM report file and experimental file here, please let me know what could be the missing steps here? Quantified_peptide_sample (1).xlsx Exp_Design_6_sample (1).txt

[trishorts]

hello. it's not clear what error you are experiencing. can you describe the problem further?

[shrivastavabhishek]

Hi @trishorts I am running "Perform Bayesian protein fold-change analysis" in the flashlfq where S1(Baseline) is the control and rest two are treatment groups, mentioned in the experimental file. I have also set the parameter as mentioned (attached screenshot). Here my concern is that I am not getting the normalised intensity values for all the samples, column number 6 onwards to 12 in the QuantifiedPeptides.tsv file. I am getting all the raw intensity values. Please have a look into attached [screenshot.]

Please let me know if you need more information about it.

[rmillikin]

Hi @shrivastavabhishek ,

Intensity normalization corrects for systematic differences in intensities between samples. It makes the median difference in intensities between two samples equal to zero. So these numbers look like they could be normalized to me. Do you mean you want log2-transformed values? Or log fold-changes between samples?

The Bayesian protein quant only computes fold-changes at the protein level, not the peptide level, also.

[shrivastavabhishek]

Hi @rmillikin Thanks for reply. Yes, I would like to get log2-transformed values, Please let me know how to acheive that. It would be a great help.

Thanks Abhishek

[rmillikin]

Ok that is actually pretty easy! Select an empty cell in Excel and type =iferror(log(A1,2),na()). Then change A1 to whatever cell(s) you want to transform. Note that you can't log a 0, you will get an error (NA value).

[shrivastavabhishek]

Hi Thanks for your reply and the excel formula @rmillikin. So, you mean to say that this log2-transformed values across the samples can not be acheived in FlashLFQ?

[rmillikin]

Currently, there isn't an option in FlashLFQ to report log intensities; you'll have to log these yourself (for example, in Excel, Perseus, etc.).

[shrivastavabhishek]

Great, Thanks for clarification.