Recently a couple of methods to derive protein relative concentrations based on peptides' covariation has been introduced to proteomics workflows. Examples include MSqRob, Diffacto, and Triqler. A common critique of such methods is that they hide interesting biology by averaging peptide's abundances over multiple proteoforms since abundances of deviating potentially proteoform specific peptides are toned down in the list of reported differential proteins. In the workshop, we will discuss and implement different measures of dispersion for individual peptides with respect to their reported protein.
Work plan
We will first discuss the theoretical advantages of different measures of a peptide's dispersion. Subsequently, we will implement a couple of such measures and compare their performance. Ideally, we form separate teams implemeting different measurement strategies.
Abstract
Recently a couple of methods to derive protein relative concentrations based on peptides' covariation has been introduced to proteomics workflows. Examples include MSqRob, Diffacto, and Triqler. A common critique of such methods is that they hide interesting biology by averaging peptide's abundances over multiple proteoforms since abundances of deviating potentially proteoform specific peptides are toned down in the list of reported differential proteins. In the workshop, we will discuss and implement different measures of dispersion for individual peptides with respect to their reported protein.
Work plan
We will first discuss the theoretical advantages of different measures of a peptide's dispersion. Subsequently, we will implement a couple of such measures and compare their performance. Ideally, we form separate teams implemeting different measurement strategies.
Technical details
The workshop will use Python, and will probably be branching out from the Diffacto (https://github.com/statisticalbiotechnology/diffacto) codebase.
We will use the iPRG2015 and the iPRG2016 datasets to characterize the different developed methods.
Contact information
Lukas Käll KTH - Royal Institute of Technology Stockholm, Sweden lukas.kall@scilifelab.se