- General Usage
Pre-processing data in Progenesis QIP
The primary method for importing data into Proteolabels is to stream the data directly from Progenesis QI for Proteomics (QIP).
To do this you must first import your data into Progenesis and perform the initial steps of map alignment, peak picking and peptide identification.
For processing data in Progenesis, please follow the user guide up to the point of peptide identification, with the following considerations:
Peak picking When performing peak picking, we strongly recommend that you select "Automatic" and "Maximum sensitivity". You may also wish to lower the filter for ions on import, as discussed in this post.
Experiment design: You do not need to consider the experimental design within Progenesis, this will be handled within Proteolabels. As such, a "Between Subject Design" is appropriate, grouping replicates within the same Condition
Peptide identification: When performing peptide identification with an MSE search or external search engine for DDA data, it is vital to ensure you identify light, (medium) and heavy peptides by using the appropriate modification settings in the search engine. Further notes are provided in the FAQ.
Selecting your search database. In general for a Proteolabels quantitation run, we recommend to use a curated reference proteome set, such as UniProt reviewed e.g. human set (~20,000 records). If you use the full UniProt set (~70,000 human sequences), it can lead to the formation of excessively large protein groups, and lots of conflicted peptides across groups. Further discussion from UniProt about the contents of the human databases can be accessed here.
Searching with Mascot: We strongly recommend to use the Quantitation settings on the search menu e.g. "SILAC K+8 R+10 [MD]" or "SILAC K+6 R+6 multiplex". We do not recommend performing two searches in Mascot with and without SILAC reagents as fixed modifications - this will cause problems in the downstream analysis. You should always make sure that identifications are filtered to ensure peptide FDR <1% in Mascot, Progenesis QIP or in Proteolabels, using the appropriate score threshold. Please consult the relevant Mascot documentation to understand how to set this threshold, as it differs depending on the Mascot version you are running.
MSE search: For MSe (data independent acquisition data), you should use the internal search engine in Progenesis QIP. You can select your quantification reagents as variable modifications, although in some scenarios, you will get optimal performance by performing the search twice (for the duplex case) - once without the quantification reagent as a modification, and once with the quantification reagent as a fixed modification.
Search with PEAKS: When searching in PEAKS, please be aware of the following considerations. Quantification reagents should be set as variable modifications, and you should filter your results (on the Summary page), to ensure < 1% FDR at the peptide-level. Before export, you need to ensure that all peptides are exported with their parent protein accession. To do this, on the Protein tab: click Show all proteins in each group, now export data into pepXML format for import into Progenesis QIP.
Refine identifications: You may choose to refine identifications within Progenesis, and then export to Proteolabels, or refine identifications within Proteolabels. In either case, it is important to ensure that only high-quality identifications are further processed to quantification.
Once you have imported your identifications you then export both the unidentified and identified features to Proteolabels directly by clicking, File > Export to Proteolabels. This menu item will only appear for users that have installed Proteolabels.
Working with fractionated data: If you have pre-fractionated your data, e.g. using 1D SDS-PAGE or OFFGEL prior to LC-MS, you can process fractionated data, and then merge in Progenesis QIP as discussed here. Once you have merged your fractions, you should then export the merged data to Proteolabels.