search

tvpham / iq

An R package to estimate relative protein abundances from ion quantification in DIA-MS-based proteomics
BSD 3-Clause "New" or "Revised" License
19 stars 9 forks source link

A quick description request for processing DIANN output #13

Open dorvashmr opened 4 months ago

dorvashmr commented 4 months ago

Hi,

First, Thank you for this really amazing package to process DIA results.

I am a bit new in using R to process and analyze data. I was wondering if you could provide a manual or example code for processing a DIANN output using most of the features that your package has. (like how it is described in detail for other software https://cran.r-project.org/web/packages/iq/vignettes/iq.html).

My other question is about the "process_wide_format" function. can you give an example of its use as well?

Cheers, Reza

chloeshep commented 4 months ago

I would really appreciate an answer to this! I am also trying to use iq in DIA-NN. Any guidance would be much appreciated.

tvpham commented 4 months ago

Hi Reza,

It is still on my todo list to write detailed examples as you suggested.

The "process_wide_format" function is to collapse multiple rows with an identical identifier into a single entry. So basically the input data is in a "wide table format".

Please feel free to put forward your questions here. I will try to answer questions here first, before any extensive documentation.

Cheers, Thang

dorvashmr commented 4 months ago

Hi again Thang,

So, if it is ok, I will put a few of my questions here and whenever you get time, I would really be grateful to read your responses.

1- If I need to use the MaxLFQ_R instead of fast_MaxLFQ, can I still use the "process_long_format"? If not, how can I do so?

2- If we need to use Top3 for a certain experiment that was searched with DIANN, how can I use the iq package?

3- in processing a diann output, what are the different normalization methods that iq package provides?

4- (this question is more on the curiosity side) I noticed that the "process_wide_format" has an argument/input parameter for "method" and it can be set to MaxLFQ or Top3 etc.. Is that an option to process DIANN report.tsv to acheive a pg_matrix?

Cheers, Reza

m-pauper commented 4 months ago

I am going to hitch on this thread and take the opportunity to also post my question.

When processing DIA-NN's report, is Iq taking into account only proteotypic peptides? If not, how could I achieve that? Using the filter_string_equal = c("Proteotypic" = "1")? In your opinion, would that be a good idea?

chloeshep commented 4 months ago

I would especially like to know the answer to: 2- If we need to use Top3 for a certain experiment that was searched with DIANN, how can I use the iq package? Thanks!

tvpham commented 4 months ago

Hi again Thang,

So, if it is ok, I will put a few of my questions here and whenever you get time, I would really be grateful to read your responses.

1- If I need to use the MaxLFQ_R instead of fast_MaxLFQ, can I still use the "process_long_format"? If not, how can I do so?

2- If we need to use Top3 for a certain experiment that was searched with DIANN, how can I use the iq package?

3- in processing a diann output, what are the different normalization methods that iq package provides?

4- (this question is more on the curiosity side) I noticed that the "process_wide_format" has an argument/input parameter for "method" and it can be set to MaxLFQ or Top3 etc.. Is that an option to process DIANN report.tsv to acheive a pg_matrix?

Cheers, Reza

Hi Reza,

  1. No at the moment, it is not possible. I will work on the package in the coming period, and I will see what is possible.
  2. You should be able to do that with the pure R version. I am trying it out. But it is quite slow.
  3. We have median normalization and no normalization.
  4. No, the process_wide_format takes as input a wide-format report and collapses multiple rows into one row. DIA-NN report.tsv is in long-format.

Thang

tvpham commented 4 months ago

I would especially like to know the answer to: 2- If we need to use Top3 for a certain experiment that was searched with DIANN, how can I use the iq package? Thanks!

I am trying out something, but it is quite slow. I will get back on this soon.

tvpham commented 4 months ago

I am going to hitch on this thread and take the opportunity to also post my question.

When processing DIA-NN's report, is Iq taking into account only proteotypic peptides? If not, how could I achieve that? Using the filter_string_equal = c("Proteotypic" = "1")? In your opinion, would that be a good idea?

Hi, please feel free to open a new issue/discussion.

Yes, filter_string_equal = c("Proteotypic" = "1") is ok. But I have not done any experiment to judge whether it is a good idea or not.