ComDim Methods for the Analysis of Multiblock Data in a Data Fusion Perspective

Cariou, Véronique; Bouveresse, Delphine Jouan-Rimbaud; Qannari, El Mostafa; Rutledge, Douglas N.

doi:10.1016/b978-0-444-63984-4.00007-7

Cited by 28 publications

(30 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In metabolomics, which is actually one of the most rapidly evolving “omics” tools measuring the small molecule composition of biofluids and tissues, representing these multiple levels of information leads to several data blocks. These latter ones, containing for the same set of samples information measured on different analytical platforms, are commonly analyzed by multiblock techniques 2–6 . Notwithstanding, a biological system can only be fully and dynamically understood when both considering its spatial and temporal dimension.…”

Section: Introductionmentioning

confidence: 99%

Three‐way clustering around latent variables approach with constraints on the configurations to facilitate interpretation

Cariou

Alexandre-Gouabau

Wilderjans

2020

Journal of Chemometrics

View full text Add to dashboard Cite

The set‐up of comprehensive studies in life sciences involving a longitudinal dimension—as appears in time‐scale metabolomics—calls for the use of dimension reduction techniques for three‐way data structures (e.g., samples by variables by time points). For this purpose, a clustering around latent variables for three‐way data approach, CLV3W, has been proposed. CLV3W aims at both partitioning the variables into nonoverlapping clusters and estimating within each cluster a rank‐one Parafac model consisting of a latent component (resp. a weighting system) associated with the first mode (resp. third mode) and a vector of loadings reflecting the degree of closeness of each variable of the second mode to its cluster. In this paper, two constrained CLV3W models are discussed. First, a nonnegativity constraint is defined implying that clusters are composed of positively correlated variables. Second, it is proposed to constrain the weighting system to be the same for all clusters. These two constraints aim at providing more parsimonious models with configurations that are easier to interpret. The appropriateness of both constraints is evaluated in a simulation study and illustrated on two case studies pertaining to sensory evaluation and metabolomics data. Regarding the first case study, CLV3W yields the identification of two consumer segments together with one common emotional pleasantness dimension associated with coffee aromas. CLV3W analysis of human preterm breast milk metabolomics data provided three clusters of lipid species that are responsible for specific functions (i.e., milk fat globules membrane‐constituents, fatty acid oxidation‐products, lipid mediators as eicosanoids and endocannabinoids).

show abstract

Section: Introductionmentioning

confidence: 99%

Three‐way clustering around latent variables approach with constraints on the configurations to facilitate interpretation

Cariou

Alexandre-Gouabau

Wilderjans

2020

Journal of Chemometrics

View full text Add to dashboard Cite

show abstract

“…Initially developed for sensory analysis, 29 ComDim or common components and specific weights analysis is a method of multiblock analysis, used for the simultaneous analysis of several matrices with different variables describing the same samples 30–32 …”

Section: Methodsmentioning

confidence: 99%

“…The global scores can be used to calculate the corresponding loadings for each normed table. The number of CCs was chosen based on the following: a plot of the cumulative sum of the saliences for each table as a function of the number of CCs, 32 the Durbin–Watson criterion, and the KMO index 34 . The latter two are mainly used for independent component analysis but can also be applied to determine the optimal number of factors in a ComDim model.…”

Section: Methodsmentioning

confidence: 99%

“…This is because scaling factors are applied to each block, so that all the blocks have the same total variance. This is done by dividing each data table by its Frobenius norm 32 …”

Section: Methodsmentioning

confidence: 99%

“…This is done by dividing each data table by its Frobenius norm. 32 As different areas of the samples were mapped, the multiblock studies were not performed on the chemical images themselves but on the mean spectra of each image. The mean NIR data were in the first block/table of data, whereas the second block contained the mean Raman spectra.…”

Section: Common Components and Specific Weights Analysis Comdimmentioning

confidence: 99%

See 2 more Smart Citations

Extraction of information about structural changes in a semisolid pharmaceutical formulation from near‐infrared and Raman images by multivariate curve resolution–alternating least squares and ComDim

Mitsutake

Neves

Rutledge

et al. 2020

Journal of Chemometrics

View full text Add to dashboard Cite

Solid dispersions are an interesting option to improve the solubility of Class II (high permeability, low water solubility) drugs of the Biopharmaceutical Classification System without the use of organic solvents. However, structural changes (polymorphism) may be present in both active pharmaceutical ingredients (API) and excipients, which require evaluation during pharmaceutical development. Thus, the aim of this work was to demonstrate the feasibility of using Raman and near-infrared (NIR) mapping associated with multivariate curve resolution-alternating least squares (MCR-ALS) and common components and specific weights analysis (CCSWA or ComDim) chemometric methods to assess the solid dispersions of atorvastatin calcium in Gelucire ® 48/16 under controlled (3.0 ± 1.0 C, sealed flask) and accelerated (33.0 ± 1.0 C, 75% of relative humidity, open flask) aging conditions. MCR-ALS allowed the identification of the amorphous and crystalline fractions of the drug within the solid dispersion, which is not easily achieved by traditional techniques such as differential scanning calorimetry and X-ray diffraction, especially for semisolid formulations. ComDim results indicated that Raman seemed more sensitive to the presence of the API, whereas NIR was found to be more sensitive to alterations in the spectra of the excipient. The use of the ComDim method is not yet widespread in the pharmaceutical area; therefore, this paper shows its potential to support pharmaceutical development in preformulation stages for stability studies.

show abstract

Chemometrics Applied to Nuclear Magnetic Resonance Analysis

Beavis

Rusilowicz

Donarski

et al. 2019

Encyclopedia of Analytical Chemistry

View full text Add to dashboard Cite

Nuclear magnetic resonance (NMR) spectroscopy provides information about the physical and chemical properties of atoms within a sample and has become an essential tool in metabolomics. Although less sensitive than mass spectrometry, NMR spectroscopy is highly reproducible and gives wider coverage in comparison to other techniques used for the analysis of complex mixtures. Recent advances in technology, advances in methodology, and increased computer power have increased the need to develop mathematical and statistical methods to analyze and interpret the complex datasets acquired. This has resulted in developments in the field of chemical informatics known as chemometrics . In contrast to targeted analyses, chemometric approaches do not initially attempt to identify particular metabolites. Instead, statistical techniques and pattern recognition methods are used to identify spectral features that show consistent trends or provide discrimination between classes. These features can provide a fingerprint for a metabolic state to be used, for example, for disease diagnostics, and individual features of interest can then be related to specific compounds and metabolic pathways using database searches. NMR‐based chemometric methods are now used in a wide range of applications including clinical diagnostics, food science and traceability, process control, monitoring genetic modification, and predicting the side effects of pharmaceuticals and their environmental effects.

show abstract

ComDim Methods for the Analysis of Multiblock Data in a Data Fusion Perspective

Cited by 28 publications

References 24 publications

Three‐way clustering around latent variables approach with constraints on the configurations to facilitate interpretation

Three‐way clustering around latent variables approach with constraints on the configurations to facilitate interpretation

Extraction of information about structural changes in a semisolid pharmaceutical formulation from near‐infrared and Raman images by multivariate curve resolution–alternating least squares and ComDim

Chemometrics Applied to Nuclear Magnetic Resonance Analysis

Contact Info

Product

Resources

About