Peak quality score

support
Peak quality score Zac  2019-08-27 09:34
 

I am looking at ways to automate the flagging of problem peaks for manual inspection. It would require some form of peak quality score. Wondering if anyone has tried to tackle this before? Or if there is way to get this indirectly in skyline, possibly from difference between a transformed and untransformed peak shape.

 
 
shadieshghi responded:  2019-08-27 13:42

Hi Zac,

There's an R package named "targetedmsqc" that does what you're describing. It calculates a number of metrics describing peak quality (symmetry, peak shape similarity, jaggedness, modality, co-elution, etc ...) and outputs an ensemble score based on the totality of the features for each transition. The current version uses Skyline reports as input and generates a csv file with columns for features and quality score. We are working on a version that allows for integration of tMSQC output as an annotation file into Skyline and hoping to release it within the next month. Please see this publication for more details if you're interested: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6173846/

The package is available here along with instructions on how to use it: https://github.com/shadieshghi/TargetedMSQC

Best,
Shadi

 
Zac responded:  2019-08-28 08:53

Thanks! Read the paper and will be trying out the package. Great list of quality features, including an assessment of peak 'jaggedness'