Systematic evaluation of normalization methods for glycomics data based on performance of network inference

Benedetti, Elisa; Gerstner, Nathalie; Pučić‐Baković, Maja; Keser, Toma; Reiding, Karli R.; Ruhaak, L. Renee; Štambuk, Tamara; Selman, Maurice H. J.; Rudan, Igor; Polašek, Ozren; Hayward, Caroline; Beekman, Marian; Slagboom, Eline; Wuhrer, Manfred; Dunlop, Malcolm; Lauc, Gordan; Krumsiek, Jan

doi:10.1101/814244

Cited by 1 publication

(2 citation statements)

References 48 publications

(44 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The challenging aspect of this sample set was introduced by the inconsistent instrument response; in the PCA plot, for example, the samples that partition on the left side of the plot generally had higher ion counts than the samples on the right side of the plot. It should be noted that when these data were classified, they had already been normalized using the standard approach in the field of reporting the glycan abundances as a fraction of 100% of the total glycan abundance [ 10 , 24 ]. So this case represents an example where the normalization strategies already in use in the field are insufficient to provide good classification outcomes, even though the problem with classification is mainly related to inconsistent ion signal.…”

Section: Resultsmentioning

confidence: 99%

“…To date the problem of signal variability, with respect to instrumentation, has been addressed most often by focusing on normalization methods. In depth studies of the best normalization practices for various sub-disciplines of mass spectrometry have been conducted, and guidelines are available for the fields of metabolomics [ 7 ], proteomics [ 8 ], glycomics [ 9 , 10 ], and imaging [ 11 ]. Any given normalization method has strengths and weaknesses, and no single method is optimal in all cases.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

The local-balanced model for improved machine learning outcomes on mass spectrometry data sets and other instrumental data

2021

View full text Add to dashboard Cite

One unifying challenge when classifying biological samples with mass spectrometry data is overcoming the obstacle of sample-to-sample variability so that differences between groups, such as between a healthy set and a disease set, can be identified. Similarly, when the same sample is re-analyzed under identical conditions, instrument signals can fluctuate by more than 10%. This signal inconsistency imposes difficulties in identifying subtle differences across a set of samples, and it weakens the mass spectrometrist’s ability to effectively leverage data in domains as diverse as proteomics, metabolomics, glycomics, and imaging. We selected challenging data sets in the fields of glycomics, mass spectrometry imaging, and bacterial typing to study the problem of within-group signal variability and adapted a 30 year old statistical approach to address the problem. The solution, “local-balanced model,” relies on using balanced subsets of training data to classify test samples. This analysis strategy was assessed on ESI-MS data of IgG-based glycopeptides and MALDI-MS imaging data of endogenous lipids, and MALDI-MS data of bacterial proteins. Two preliminary examples on non-mass spectrometry data sets are also included to show the potential generality of the method outside the field of MS analysis. We demonstrate that this approach is superior to simple normalization methods, generalizable to multiple mass spectrometry domains, and potentially appropriate in fields as diverse as physics and satellite imaging. In some cases, improvements in classification can be dramatic, with accuracy escalating from 60% with normalization alone to over 90% with the additional development described herein.

show abstract

Section: Resultsmentioning

confidence: 99%