Coming from a background where internal standards is an absolute must for quantification, I wonder about spectral counting in proteomics. How can this work properly, considering ion suppression effects and the immense complexity of a proteomics sample?
It is a rough-and-ready method and has a relatively high variance even between tech replicates, especially where lower count values are concerned - these issues are covered in a number of reviews as pointed out by the others. That much said - it comes for free (directly from the protein ID runs) and can help differentiate larger abundance changes (in our hands >5x), and is a good technique for AP-MS where specific pulldowns show strong enrichment vs. controls. For complex samples, there are way more powerful and reproducible techniques out there by now.
Hi Steven,
Please check:
http://www.ncbi.nlm.nih.gov/pubmed/15253663
http://www.ncbi.nlm.nih.gov/pubmed/15253663
And probably related to the normalisation strategy Nazrul is using:
http://www.ncbi.nlm.nih.gov/pubmed/16944946
http://www.ncbi.nlm.nih.gov/pubmed/20166708
Best, Tiago
The data acquisition method ( Data dependent or independent acquisition) employed in MS also affects the reliability of label free quantitation methods including spectral counting.
http://www.ncbi.nlm.nih.gov/pubmed/25348682
It is a rough-and-ready method and has a relatively high variance even between tech replicates, especially where lower count values are concerned - these issues are covered in a number of reviews as pointed out by the others. That much said - it comes for free (directly from the protein ID runs) and can help differentiate larger abundance changes (in our hands >5x), and is a good technique for AP-MS where specific pulldowns show strong enrichment vs. controls. For complex samples, there are way more powerful and reproducible techniques out there by now.
Hi Steven
Spectral counting is another form of relative quantitation and limited to fairly large changes between comparative samples, like treated or untreated. It is not a method to get absolute quantitation, so the question of trust is completely overshadowed by how well you work and how good your controls are. If there are more than one unique peptide per protein the difference of each unique peptide should be about the same and these do improve the probability of more accurately determining the differences, but care should be taken at all steps.
Bottom line is it does work but be careful about how you apply the results.
Hi Everyone, thanks for the input! Duncan, I appreciate the point of checking several peptides, that can help reduce the fear of a suppression getting in the way. But I guess users typically confirm their results with a complementary follow-up study with more robustness, I suppose?
- Badges
- Science topic
- Similar topics
- Chemistry
- General Biochemistry
- Proteomics