Effect of quality characteristics of single sample preparation steps in the precision and coverage of proteomic studies—A review

Proteomic profiling and biomarker search are analytical tools as many other. Nevertheless, in the proteomic discovery phase considerable sample fractionation is inevitable before readout. Since these procedures are of notable complexity, proteomic tools need in particular analytical quality validation standards as prevail for other analytical methods. With acceptance of the rule of error propagation the values of imprecision and yield of each preparation step determine overall reproducibility and therewith information harvest of a propagated method series. Thereto, we examined recent proteomic reports with reproducibility data and with parallelization, and automation approaches. Based on the data available from literature it is highly probable, that at least a part of current proteomic platforms actually suffer from high technical variance.The general need for quantification and the impact of precision and recovery levels with each step of typical multi-step workflows are illustrated. All sample preparation approaches that maximize yields (percent recoveries) and minimize overall imprecision should be suitable to improve reliability of biomarker search. This has to be realized by technical innovations that (a) minimize the imprecision of each serial sample preparation step, that (b) minimize the number of serial sample preparation steps, and that (c) parallelize all procedures with a maximum of automation.

Figure optionsDownload as PowerPoint slideHighlights
► In contrast to usual analytics exorbitant numbers of steps are combined in proteomics.
► Overall precision and yield are virtually impossible to be determined experimentally.
► We estimated overall bias using published values of various singular procedures.
► Unexpectedly, calculated overall bias is in the range of aberrations of many candidate biomarkers.
► Proposals to improve precision and yield include hyphenation, parallelization and automation.