How to distinguish healthy from diseased? Classification strategy for mass spectrometry‐based clinical proteomics

Abstract

SELDI-TOF-MS is rapidly gaining popularity as a screening tool for clinical applications of proteomics. Application of adequate statistical techniques in all the stages from measurement to information is obligatory. One of the statistical methods often used in proteomics is classification: the assignment of subjects to discrete categories, for example healthy or diseased. Lately, many new classification methods have been developed, often specifically for the analysis of X-omics data. For proteomics studies a good strategy for evaluating classification results is of prime importance, because usually the number of objects will be small and it would be wasteful to set aside part of these as a 'mere' test set. The present paper offers such a strategy in the form of a protocol which can be used for choosing among different statistical classification methods and obtaining figures of merit of their performance. This paper also illustrates the usefulness of proteomics in a clinical setting, serum samples from Gaucher disease patients, when used in combination with an appropriate classification method.

Characterization of the lipid profile post mortem for Type-2 diabetes in human brain and plasma of the elderly with LCMS-techniques: a descriptive approach of diabetic encephalopathy

“…Via our descriptive approach we can conclude that for biochemical lipid composition for T2DM patients the brain is very stable and that mental disorders more have to be found in the neural functioning (conduction and/or transmission) resulting in slowing down brain functioning (38,39) and glycaemic control than in the biochemical lipid composition. Our major conclusion is that diabetes and its treatment are more associated with glycaemic control than structural disturbances (lipid composition) in the brain.…”

“…This is a supervised method forcing group separation based on the principle that PCDA is prone to overfitting (false positive models). PCDA is a powerful tool to identify and maximize differences between pre-defined groups in data sets with a large number of variables [39].…”

Statistical data processing in clinical proteomics

“…For classification of MS data of Gaucher disease, Hendriks et al applied six classification methods [67]. The most successful were SVM, penalized logistic regression and PCDA.…”

“…Alternatively, the rules can all be constructed with the same classification method, for example ANN [61]. Diversity of the rules can then be introduced by resampling the subjects with cross-validation [62], bootstrapping [61,[63][64][65], and boosting [66,67]. A combination of bagging and boosting is used by Dettling in BagBoosting, where in each boosting step a bagged classifier is constructed [68].…”

Biomarker Characterization by MALDI–TOF/MS

“…Conventionally, MALDI-TOF/MS combined with two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) has been the main analytical method for identifying proteins in bottom-up proteomics [123][124][125]. In clinical proteomics, this approach has been used to evaluate the expression of specific proteins or peptides in biological samples between controls and patients [126][127][128]. However, tedious sample treatment required for 2D-PAGE makes this approach impractical in clinical analysis which always needs high-throughput capability to digest a large number of the samples.…”