Why do protein groups sometimes need to be rebuilt when I include/exclude some identifications?

Progenesis QI for proteomics has the option to simplify its protein list by grouping together proteins which have been identified by similar sets of peptide sequences. If you edit peptide identifications with this grouping option selected, a message may appear telling you that you need to rebuild protein groupings. This is because the edits you have made have changed the way proteins should be grouped together.

How does protein grouping work?

A protein will be grouped with another if its identified peptide sequences are a subset of the identified sequences of the other. For example if we have a set of proteins each with a number of identified sequences.

Protein2 and Protein3 will both be grouped under Protein1 as the sequences for Protein2 and Protein3 are both subsets of Protein1.

Protein4 shares 2 sequences with Protein1 but will not be grouped with it as it has one sequence 'AFNGLLR' which does not appear in Protein1's sequence list. This means there is independent evidence that Protein4 is present in the sample and so it will remain in the protein list.

How can conflict resolution affect this?

During conflict resolution you can 'un-tick' peptide identifications to indicate that the software should not consider that identification to be reliable. So, in the example above, you could decide that the identification of sequence 'AFNGLLR' of Protein4 is unreliable and un-tick it. In this case the proteins would now look like:

Protien4's peptides are now just a subset of Protein1's. As there is now no independent evidence for the existence of Protein4 it should be removed from the protein list and grouped under Protein1. Progenesis QI for proteomics warns you of this by showing this message.

When this happens press 'Rebuild' to rebuild the protein groups taking into account the changes to the included identifications. Protein4 will then disappear from the protein list.