Eg Actn1 detected twice, giving to very different Log2FC. Do I use the sequence coverage %?
Dear Matilde,
your question is not easily answered. First it is essential to know which paramters you used for your MQ search, especially which peptides were used for quantification (only unique or unique and razor) and which FASTA was used (SwissProt only SwissProtTrembl which is also called "reference proteome" or a fasta with isoforms). Most likely the double quantification of Actn1 stems from protein grouping, means that the identified peptides for this protein were not unique, and thus it was put in a protein group with e.g. isoforms or Actn1 has more than one accession in your FASTA. It would make sense for you to look at the peptide level to see all quantified peptides of Actn1, this could give you more information on its true behaviour and of course the protein groups (accessions) associated with your two quantified Actn1, to see which other proteins were grouped with it. Hope this helps!
BW,
Britta
Dear Britta,
Thanks a lot for taking your time to answer my question.
I am quite new to MQ, so I am not familiar with all the settings that were set for this analysis. I have attached a file, with raw data for MQ where the first tab "raw data" , column G "gene names" shows the duplicates in red. In this case eg. Anxa4 has the same peptide sequence in both cases column (BD)?
Dear Mathilde,
I would suggest to increase the "min unique peptides" to 2 in MQ (Global Parameters/Identification tab) to be on the sure side. Also, I would not trust the LFQ value solely based on one unique peptide. Sure, the number of IDs in your table will be much smaller but the remaining once will be more reliable.
Regarding the mentioned Annexins, the second Annexin entry seems to be a truncated version with a 100 % similarity of the first 176 N-terminal AAs but the remaining sequence of 20 AAs of the C-terminus is completely different from the Majority Protein from the first Annexin entry. This is also the region where most of the sequence of the one unique peptide (from the second Annexin) is from.
Anyway, usually after statistics (including FDR) the remaining significant once will help to filter out this problems. In other words, LFQ data based on one peptide only will show more variances between the replicates and this will lead to higher p-values.
Best,
Murat
- Badges
- Science method