Generalized simultaneous component analysis of binary and quantitative data. (5th November 2020)

• Record Type: Journal Article
• Authors: Song, Yipeng; Westerhuis, Johan A.; Aben, Nanne; Wessels, Lodewyk F. A.; Groenen, Patrick J. F.; Smilde, Age K.
• Is Part Of: Journal of chemometrics. Volume 35:Number 3(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-11-05
• Subjects: binary data -- component model -- concave penalty -- data fusion; Chemistry -- Mathematics -- Periodicals; Chemistry -- Statistical methods -- Periodicals; 542.85
• DOI: 10.1002/cem.3312
• Languages: English
• ISSNs: 0886-9383

Abstract

Abstract: In the current era of systems biology research, there is a need for the integrative analysis of binary and quantitative genomics data sets measured on the same objects. One standard tool of exploring the underlying dependence structure present in multiple quantitative data sets is the simultaneous component analysis (SCA) model. However, it does not have any provisions when a part of the data are binary. To this end, we propose the generalized SCA (GSCA) model, which takes into account the distinct mathematical properties of binary and quantitative measurements in the maximum likelihood framework. Like in the SCA model, a common low‐dimensional subspace is assumed to represent the shared information between these two distinct types of measurements. To achieve a low rank solution, we propose to use a concave variant of the nuclear norm penalty. An efficient majorization algorithm is developed to fit this model with different concave penalties. Realistic simulations (low signal‐to‐noise ratio and highly imbalanced binary data) are used to evaluate the performance of the proposed model in recovering the underlying structure. Also, a missing value based cross‐validation procedure is implemented for model selection. We illustrate the usefulness of the GSCA model for exploratory data analysis of quantitative gene expression and binary copy number aberration measurements obtained from the GDSC1000 data sets. Abstract : For the integrative analysis of binary and quantitative data sets, we propose the generalized simultaneous component analysis model. A common low‐dimensional subspace is assumed to represent the shared information between these two distinct types of measurements. To achieve a low‐rank solution, we propose to use a concave variant of the nuclear norm penalty. Efficient algorithms are developed. The model is validated by realistic simulations, and its usefulness is illustrated by the analysis of the GDSC1000 data sets.