-
### Description of feature
I tested identification workflow for timsTOF dataset in last week. The first step is to execute the tdf2mzml module and then run the same analyses as for the other data (C…
-
Hello, I used DIA-NN's library free and deep learning mode to identify Astral DDA data and have the following questions about the results:
1. Some PSMs do not have corresponding peaks of the identifi…
-
Hello. I'm trying to use PeptideShaker to process identification files obtained by SearchGUI using several search engines (Comet, OMSSA, Tide, X! Tandem and Sage).
PeptideShaker encounters the follow…
-
Dear developers,
I am encountering an issue with peptide-shaker, which I paste here:
# ------------------------------------------------------------------
# PeptideShaker 2.2.25 Report File
#
…
-
Dear Fragpipe team,
We are comparing identification and quantitation numbers in spectral data acquired in DIA mode from TimsTOF Ultra and Thermo Astral.
Samples are N-terminally enriched extract…
-
Hello,
we have generated an HLA-peptides library using PeptDeep-HLA for the prediction of peptides and alphapeptdeep for library generation and conversion to .tsv for compatibility with DIANN. Then…
-
### PXD001819 Analysis
Currently, we have a workflow that can perform peptide identification using: `-> ms2rescore -> SNR + spectrum properties -> percolator`
Here the results can be found (http…
-
Hi,
We are performing glycopeptide identification and quantification across many samples. The data in the "assigned modifications" column of the "combined_modified_peptide.tsv" file gives only the …
-
Hi HiTMaP team,
Currently I have been running through your mouse brain example data. However, I encountered an error:
## Codes I used:
```
fastafile
-
In recent data, we encountered a problem with peptides, that are so stable they do not undergo exchange in both states. For our own use, we call them "płaszczaki". The issue is that the maximal exchan…