Known Knowns, Known Unknowns, and Unknown Unknowns: Coverage in MS Experiments.

Area of Science:

• Ecology
• Biochemistry
• Computational Biology

Background:

• Mass spectrometry experiments identify proteins but may miss some.
• Estimating the total number of proteins (proteome cardinality) is crucial.
• Ecological capture-recapture models are effective for population size estimation.

Purpose of the Study:

• To adapt ecological capture-recapture models for analyzing mass spectrometry data.
• To estimate the total number of protein identifications in proteomic experiments.
• To explore the broader applicability of these models for estimating missing data.

Main Methods:

• Implementation of a closed-population capture-recapture model with time-varying and heterogeneous capture probabilities.
• Utilizing the R package Rcapture for model fitting and proteome cardinality estimation.
• Alternative model fitting using general linear models in software like Matlab.

Main Results:

• The capture-recapture model successfully models protein identifications across mass spectrometry cycles.
• Rcapture provides straightforward estimates of proteome size.
• The methodology is adaptable for estimating missing observations in various experimental contexts.

Conclusions:

• Capture-recapture models offer a robust framework for proteomic data analysis.
• These models enhance the estimation of proteome cardinality from mass spectrometry.
• The approach has potential for broader applications in estimating missing experimental data.