Scale-Invariant Biomarker Discovery in Urine and Plasma Metabolite Fingerprints

Zacharias, Helena U.; Rehberg, Thorsten; Mehrl, Sebastian; Richtmann, Daniel; Wettig, Tilo; Oefner, Peter J.; Spang, Rainer; Gronwald, Wolfram; Altenbuchinger, Michael

doi:10.1021/acs.jproteome.7b00325

Cited by 17 publications

(26 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A particular type of information, which is assumed to be affected, will be neglected in subsequent modelling processes. This work is related somehow to work by the group of Rainer Spang on zero-sum regression [11,12]; in fact, our classifier C con corresponds to this concept class. Here, we extend and generalize this approach and also embed it into the PAC learning framework.…”

Section: Discussionmentioning

confidence: 89%

See 1 more Smart Citation

Constraining classifiers in molecular analysis: invariance and robustness

Lausser

Szekely

Klimmek

et al. 2020

J. R. Soc. Interface.

View full text Add to dashboard Cite

Analysing molecular profiles requires the selection of classification models that can cope with the high dimensionality and variability of these data. Also, improper reference point choice and scaling pose additional challenges. Often model selection is somewhat guided by ad hoc simulations rather than by sophisticated considerations on the properties of a categorization model. Here, we derive and report four linked linear concept classes/models with distinct invariance properties for high-dimensional molecular classification. We can further show that these concept classes also form a half-order of complexity classes in terms of Vapnik–Chervonenkis dimensions, which also implies increased generalization abilities. We implemented support vector machines with these properties. Surprisingly, we were able to attain comparable or even superior generalization abilities to the standard linear one on the 27 investigated RNA-Seq and microarray datasets. Our results indicate that a priori chosen invariant models can replace ad hoc robustness analysis by interpretable and theoretically guaranteed properties in molecular categorization.

show abstract

Section: Discussionmentioning

confidence: 89%

“…In the following, we focus on incorporating invariances into classification models [10]. Other approaches focus on regression applications [11,12]. That is the classification model and its predictions should not be affected by a specific data transformation.…”

Section: Introductionmentioning

confidence: 99%

Constraining classifiers in molecular analysis: invariance and robustness

Lausser

Szekely

Klimmek

et al. 2020

J. R. Soc. Interface.

View full text Add to dashboard Cite

show abstract

“…However, the results of this data analysis strategy exhibit a distinct dependency on the a priori chosen normalization method, as investigated by [18]. Figure 2 illustrates these observations for a ttest analysis of urinary NMR fingerprints of acute kidney injury (AKI) vs. healthy patients.…”

Section: Supervised Data Analysismentioning

confidence: 98%

“…In this case, it is possible to learn the scaling parameters on a subset of non-regulated features only [17]. However, it is important to note that all of the different scaling strategies impact the following analysis steps such as screening for differential metabolites or multivariate metabolic signatures [18][19][20][21]. Following data scaling it is often necessary to adjust the variance of the data.…”

Section: Spectra Processing and Data Preprocessingmentioning

confidence: 99%

Data Normalization in NMR-based Metabolomics

Zacharias¹,

Altenbuchinger²,

Gronwald³

2018

Preprint

Self Cite

View full text Add to dashboard Cite

The aim of this article is to summarize recent bioinformatic and statistical developments applicable to NMR-based metabolomics. Extracting relevant information from large multivariate datasets by statistical data analysis strategies may be of considerable complexity. Typical tasks comprise for example classification of specimens, identification of differentially produced metabolites, and estimation of fold changes. In this context it is of prime importance to minimize contributions from unwanted biases and experimental variance prior to these analyses. This is the goal of data normalization. Therefore, special emphasize is given to different data normalization strategies. In the first part, we will discuss the requirements and the pros and cons for a variety of commonly applied strategies. In the second part, we will concentrate on possible solutions in case that the requirements for the standard strategies are not fulfilled. In the last part, very recent developments will be discussed that allow reliable estimation of metabolic signatures for sample classification without prior data normalization. In this contribution special emphasis will be given to techniques that have worked well in our hands.

show abstract

“…zeroSum is an R package allows the transfer of metabolic signatures across labs . zeroSum identifies the same biomarkers regardless of the scaling method.…”

Section: Miscellaneous Tools Of Interestmentioning

confidence: 99%

New tools and resources in metabolomics: 2016–2017

Misra¹

2018

Electrophoresis

View full text Add to dashboard Cite

Rapid advances in mass spectrometry (MS) and nuclear magnetic resonance (NMR)-based platforms for metabolomics have led to an upsurge of data every single year. Newer high-throughput platforms, hyphenated technologies, miniaturization, and tool kits in data acquisition efforts in metabolomics have led to additional challenges in metabolomics data pre-processing, analysis, interpretation, and integration. Thanks to the informatics, statistics, and computational community, new resources continue to develop for metabolomics researchers. The purpose of this review is to provide a summary of the metabolomics tools, software, and databases that were developed or improved during 2016-2017, thus, enabling readers, developers, and researchers access to a succinct but thorough list of resources for further improvisation, implementation, and application in due course of time.

show abstract

Scale-Invariant Biomarker Discovery in Urine and Plasma Metabolite Fingerprints

Cited by 17 publications

References 40 publications

Constraining classifiers in molecular analysis: invariance and robustness

Constraining classifiers in molecular analysis: invariance and robustness

Data Normalization in NMR-based Metabolomics

New tools and resources in metabolomics: 2016–2017

Contact Info

Product

Resources

About