Relationship between the lipid composition of maternal plasma and infant plasma through breast milk

Introduction This study was motivated by the report that infant development correlates with particular lipids in infant plasma. Objective The hypothesis was that the abundance of these candidate biomarkers is influenced by the dietary intake of the infant. Methods A cohort of 30 exclusively-breastfeeding mother–infant pairs from a small region of West Africa was used for this observational study. Plasma and milk from the mother and plasma from her infant were collected within 24 h, 3 months post partum. The lipid, sterol and glyceride composition was surveyed using direct infusion MS in positive and negative ion modes. Analysis employed a combination of univariate and multivariate tests. Results The lipid profiles of mother and infant plasma samples are similar but distinguishable, and both are distinct from milk. Phosphatidylcholines (PC), cholesteryl esters (CEs) and cholesterol were more abundant in mothers with respect to their infants, e.g. PC(34:1) was 5.66% in mothers but 3.61% in infants (p = 3.60 × 10−10), CE(18:2) was 8.05% in mothers but 5.18% in infants (p = 1.37 × 10−11) whilst TGs were lower in mothers with respect to their infants, e.g. TG(52:2) was 2.74% in mothers and 4.23% in infants (p = 1.63 × 10−05). A latent structure model showed that four lipids in infant plasma previously shown to be biomarkers clustered with cholesteryl esters in the maternal circulation. Conclusion This study found evidence that the abundance of individual lipid isoforms associated with infant development are associated with the abundance of individual molecular species in the mother’s circulation.

A Bayesian two-way latent structure model for genomic data integration reveals few pan-genomic cluster subtypes in a breast cancer cohort

Motivation Unsupervised clustering is important in disease subtyping, among having other genomic applications. As genomic data has become more multifaceted, how to cluster across data sources for more precise subtyping is an ever more important area of research. Many of the methods proposed so far, including iCluster and Cluster of Cluster Assignments (COCAs), make an unreasonable assumption of a common clustering across all data sources, and those that do not are fewer and tend to be computationally intensive. Results We propose a Bayesian parametric model for integrative, unsupervised clustering across data sources. In our two-way latent structure model, samples are clustered in relation to each specific data source, distinguishing it from methods like COCAs and iCluster, but cluster labels have across-dataset meaning, allowing cluster information to be shared between data sources. A common scaling across data sources is not required, and inference is obtained by a Gibbs Sampler, which we improve with a warm start strategy and modified density functions to robustify and speed convergence. Posterior interpretation allows for inference on common clusterings occurring among subsets of data sources. An interesting statistical formulation of the model results in sampling from closed-form posteriors despite incorporation of a complex latent structure. We fit the model with Gaussian and more general densities, which influences the degree of across-dataset cluster label sharing. Uniquely among integrative clustering models, our formulation makes no nestedness assumptions of samples across data sources so that a sample missing data from one genomic source can be clustered according to its existing data sources. We apply our model to a Norwegian breast cancer cohort of ductal carcinoma in situ and invasive tumors, comprised of somatic copy-number alteration, methylation and expression datasets. We find enrichment in the Her2 subtype and ductal carcinoma among those observations exhibiting greater cluster correspondence across expression and CNA data. In general, there are few pan-genomic clusterings, suggesting that models assuming a common clustering across genomic data sources might yield misleading results. Availability and implementation The model is implemented in an R package called twl (‘two-way latent’), available on CRAN. Data for analysis are available within the R package. Supplementary information Supplementary data are available at Bioinformatics online.

A Bayesian Two-Way Latent Structure Model for Genomic Data Integration Reveals Few Pan-Genomic Cluster Subtypes in a Breast Cancer Cohort

MotivationUnsupervised clustering is important in disease subtyping, among having other genomic applications. As genomic data has become more multifaceted, how to cluster across data sources for more precise subtyping is an ever more important area of research. Many of the methods proposed so far, including iCluster and Cluster of Cluster Assignments, make an unreasonble assumption of a common clustering across all data sources, and those that do not are fewer and tend to be computationally intensive.ResultsWe propose a Bayesian parametric model for integrative, unsupervised clustering across data sources. In our two-way latent structure model, samples are clustered in relation to each specific data source, distinguishing it from methods like Cluster of Cluster Assignments and iCluster, but cluster labels have across-dataset meaning, allowing cluster information to be shared between data sources. A common scaling across data sources is not required, and inference is obtained by a Gibbs Sampler, which we improve with a warm start strategy and modified density functions to robustify and speed convergence. Posterior interpretation allows for inference on common clusterings occurring among subsets of data sources. An interesting statistical formulation of the model results in sampling from closed-form posteriors despite incorporation of a complex latent structure. We fit the model with Gaussian and more general densities, which influences the degree of across-dataset cluster label sharing. Uniquely among integrative clustering models, our formulation makes no nestedness assumptions of samples across data sources so that a sample missing data from one genomic source can be clustered according to its existing data sources.We apply our model to a Norwegian breast cancer cohort of ductal carcinoma in-situ and invasive tumors, comprised of somatic copy-number alteration, methylation and expression datasets. We find enrichment in the Her2 subtype and ductal carcinoma among those observations exhibiting greater cluster correspondence across expression and CNA data. In general, there are few pan-genomic clusterings, suggesting that models assuming a common clustering across genomic data sources might yield misleading results.Implementation and AvailabilityThe model is implemented in an R package called twl (“two-way latent”), available on CRAN. Data for analysis is available within the R [email protected] MaterialAppendices are available online and include additional Breast Cancer subtyping analysis and model runs, comparison with leading integrative clustering methods, fully general statistical formulation and description of improvements of the Gibbs sampler.

Revisiting the pathways to retirement: a latent structure model of the dynamics of transition from work to retirement

ABSTRACTA dynamic latent structure model of the work–retirement transition process was identified, focusing on transitions of work and retirement status for men and women aged 51–74 years. Using the Health and Retirement Study data (1998–2004), latent transition analysis was used to identify a best fitting model capturing work–retirement statuses in four samples defined by age and sex. The prevalence of each status was described and the dynamic transition probabilities within the latent structure were examined. Using multinomial logistic regression, socio-demographic, health, family and occupational factors were assessed to determine how each was related to the likelihood of occupying a specific latent status at baseline. Results showed that study respondents were classified into distinct groups: full retiree, partial retiree or part-time worker, full-time worker, work-disabled or home-maker. The prevalence of full retiree status increased, while the prevalence for full-time worker status decreased over time for both men and women. Membership rates in the work-disabled and partial retiree status were generally consistent, with decreased probabilities of the work-disabled status in the older age groups and increased probabilities of partial retirees among younger men. Our findings indicated that many older Americans experience multiple transitions on the pathway to retirement. Future research on late-life labour-force transitions should evaluate the impact of the recent Great Recession and examine the role of larger socio-economic contexts.