Advantages and challenges of label-free quantitative LC-MS
Label-free LC-MS gives your quantitative proteomics advantages that are orders of magnitude better than other techniques, including labelling 1,2. But analysis of label-free data is challenging, particularly if quantification relies on detection and identification of unique peptides for each protein of interest.
|
Source: A Comparison of Labelling and Label-Free Mass Spectrometry-Based Proteomics Approaches, Vibhuti J. Patel, Konstantinos Thalassinos, Susan E. Slade, Joanne B. Connolly, Andrew Crombie, J. Colin Murrell and James H. Scrivens. J. Proteome Res., 2009, 8 (7), pp 3752–3759. May 12, 2009
Progenesis LC-MS takes a different approach by quickly, reliably and objectively quantifying what’s changing before automatically bringing together quantitative data with qualitative information and then report changes at the protein level.
How does Progenesis LC-MS achieve this?
Peak modelling. Online LC-MS can generate very large data sets. For this reason data is often centroided, resulting in a loss of valuable information. To overcome this and allow the handling of large numbers of samples, we’ve developed an intelligent peak-modelling algorithm that can reduce data files by an order of magnitude. Using a wavelet based approach, peaks are identified and peak models created that retain all relevant quantification and positional information.
LC-MS run alignment and 100% matching. The ability to combine data from multiple LC-MS runs is required for comparative expression profiling studies. This enables the comparison of different experimental conditions using significant levels of biological and technical replicates. To combine and compare results from different run the LC-MS data Progenesis LC-MS aligns them to compensate for the positional bias introduced by the LC separation technique. The result is increased reliability and reproducibility of results.
Alignment is automated and based on paired feature detection at the LC-MS level; it doesn’t reduce the data to the total ion chromatogram. This is followed by regression analysis in peptides retention time and m/z to produce an alignment grid used to accurately overlay the data.
LC-MS runs represented as m/z vs. RT overlaid (A) unaligned runs (B) aligned runs
Total Ion Chromatogram view (A) unaligned runs (B) aligned runs
To ensure consistent peptide detection and matching across all data files, an aggregate data set is created from the aligned runs. This contains all peak information from all sample files, allowing the detection of a single peptide map which is applied to each sample. This powerful approach ensures 100% matching of peptide features and enables you to apply multivariate statistical tools to explore peptide data and measure differential analysis. Peak data is combined in such a way as to maintain peptide shape properties.
Peptide Detection. Using the aggregated peak data, peptides are automatically identified creating a map of all peptides on all samples. An associated charge and mass is assigned to each peptide. Detection uses isotope profiles to confidently build peptides utilising three dimensional data.
Differential Expression Analysis
Quantification of features in LC-MS runs then targeted identification using inclusion lists. You can take a simple approach by generating runs in LC-MS mode to quantify all your differentially expressed peptide ions, then quickly generate an inclusion list of only the significantly changing features for targeted LC-MS/MS runs. The MS/MS spectra you generate can be added to your existing experiment without affecting quantification. This has the added benefit of reducing the complexity of the data you need to investigate and helps you run significant numbers of replicates within a shorter time.
Quantification and identification of features within the same LC-MS/MS runs. This involves analysing LC-MS/MS runs that already contain quantification and identification data within one run. In this case after quantifying all your peptide ions and filtering down to the ones that are most significant or that you're most interested in, you automatically collate them with qualitative information from database searches of the associated MS/MS. This means you can easily put back together the complex pattern of pieces generated by LC-MS/MS in one analysis and quickly answer questions about what is happening at the protein level. As an added benefit of our analysis approach any low abundance peptide ions that fail to trigger MS/MS data can be targeted for LC-MS/MS as above.
Protein Characterisation
This protein analysis approach to characterise a simple protein mix or pure protein and the population of peptides they generate is also supported within the same product. In this case LC-MS/MS runs are used to map the protein and interesting peptide modifications within the sample. Observing peptide ions behaving differently between samples of the same protein can indicate the presence and location of protein adducts or post-translational modifications. This information can help you to define the mode of action for a protein of interest. An additional benefit of our analysis approach with its visual features mean you can explore the LC-MS data in a simple view to find any peptide ion based on its critical attributes e.g. m/z, mass, RT. From this point you can select peptide ions for further LC-MS/MS runs to build up qualitative data where this information is missing.
1 Quantitative mass spectrometry in proteomics: a critical review. Anal Bioanal Chem (2007), 389:1017-1031 Marcus Bantscheff, Markus Schirle, Gavain Sweetman, Jens Rick, Berhard Kuster
2 A Comparison of Labelling and Label-Free Mass Spectrometry-Based Proteomics Approaches, Vibhuti J. Patel, Konstantinos Thalassinos, Susan E. Slade, Joanne B. Connolly, Andrew Crombie, J. Colin Murrell and James H. Scrivens. J. Proteome Res., 2009, 8 (7), pp 3752–3759. May 12, 2009
