Multiple imputation for model checking: completed-data plots with missing and latent data

In problems with missing or latent data, a standard approach is to first impute the unobserved data, then perform all statistical analyses on the completed dataset--corresponding to the observed data and imputed unobserved data--using standard procedures for complete-data inference. Here, we extend this approach to model checking by demonstrating the advantages of the use of completed-data model diagnostics on imputed completed datasets. The approach is set in the theoretical framework of Bayesian posterior predictive checks (but, as with missing-data imputation, our methods of missing-data model checking can also be interpreted as "predictive inference" in a non-Bayesian context). We consider the graphical diagnostics within this framework. Advantages of the completed-data approach include: (1) One can often check model fit in terms of quantities that are of key substantive interest in a natural way, which is not always possible using observed data alone. (2) In problems with missing data, checks may be devised that do not require to model the missingness or inclusion mechanism; the latter is useful for the analysis of ignorable but unknown data collection mechanisms, such as are often assumed in the analysis of sample surveys and observational studies. (3) In many problems with latent data, it is possible to check qualitative features of the model (for example, independence of two variables) that can be naturally formalized with the help of the latent data. We illustrate with several applied examples.     

Molecular phylogeny of the springhare, Pedetes capensis, based on mitochondrial DNA sequences

The phylogenetic position of the Pedetidae, represented by a single species Pedetes capensis, is controversial, reflecting in part the retention of both Hystricomorphous and Sciurognathous characteristics in this rodent. In an attempt to clarify the species evolutionary relationships, mtDNA gene sequences from 10 rodent species (representing seven families) were analyzed using phenetic, parsimony, and maximum-likelihood methods of phylogenetic inference; the rabbit, Oryctolagus cuniculus (Order Lagomorpha), and cow, Bos taurus (Order Artiodactyla), were used as outgroups. Investigation of 714 base pairs of the protein-coding cytochrome b gene indicate strong base bias at the third codon position with significant rate heterogeneity evident between the three structural domains of this gene. Similar analyses conducted on 816 base pairs of the 12S rRNA gene revealed a transversion bias in the loop sections of all taxa. The cytochrome b gene sequences proved useful in resolving associations between closely related species but failed to produce consistent tree topologies at the family level. In contrast, phylogenetic analysis of the 12S rRNA gene resulted in strong support for the clustering of Pedetidae/Heteromyidae/Geomyidae and Muridae in one clade to the exclusion of the Hystricidae/Thryonomyidae and Sciuridae, a finding which is concordant with studies of rodent fetal membranes as well as reproductive and other anatomical features.      

Flexible Modelling of the Covariance Matrix in a Linear Random Effects Model

A flexible approach is proposed for modelling the covariance matrix of a linear mixed model for longitudinal data. The method combines parametric modelling of the random effects part with flexible modelling of the serial correlation component. The approach is exemplified on weight gain data and on the evolution of height of children in their first year of life of the Jimma Infant Survival Study, an Ethiopian cohort study. The analyses show the usefulness of the approach.     

Revisiting GMOs: Are There Differences in European Consumers’ Acceptance and Valuation for Cisgenically vs Transgenically Bred Rice?

Both cisgenesis and transgenesis are plant breeding techniques that can be used to introduce new genes into plant genomes. However, transgenesis uses gene(s) from a non-plant organism or from a donor plant that is sexually incompatible with the recipient plant while cisgenesis involves the introduction of gene(s) from a crossable—sexually compatible—plant. Traditional breeding techniques could possibly achieve the same results as those from cisgenesis, but would require a much larger timeframe. Cisgenesis allows plant breeders to enhance an existing cultivar more quickly and with little to no genetic drag. The current regulation in the European Union (EU) on genetically modified organisms (GMOs) treats cisgenic plants the same as transgenic plants and both are mandatorily labeled as GMOs. This study estimates European consumers’ willingness-to-pay (WTP) for rice labeled as GM, cisgenic, with environmental benefits (which cisgenesis could provide), or any combination of these three attributes. Data were collected from 3,002 participants through an online survey administered in Belgium, France, the Netherlands, Spain and the United Kingdom in 2013. Censored regression models were used to model consumers’ WTP in each country. Model estimates highlight significant differences in WTP across countries. In all five countries, consumers are willing-to-pay a premium to avoid purchasing rice labeled as GM. In all countries except Spain, consumers have a significantly higher WTP to avoid consuming rice labeled as GM compared to rice labeled as cisgenic, suggesting that inserting genes from the plant’s own gene pool is more acceptable to consumers. Additionally, French consumers are willing-to-pay a premium for rice labeled as having environmental benefits compared to conventional rice. These findings suggest that not all GMOs are the same in consumers’ eyes and thus, from a consumer preference perspective, the differences between transgenic and cisgenic products are recommended to be reflected in GMO labeling and trade policies.     

The role of CXC chemokines in the transition of chronic inflammation to esophageal and gastric cancer

Chronic inflammation may increase the risk to develop cancer, for instance esophagitis or gastritis may lead to development of esophageal or gastric cancer, respectively. The key molecules attracting leukocytes to local inflammatory sites are chemokines. We here provide a systematic review on the impact of CXC chemokines (binding the receptors CXCR1, CXCR2, CXCR3 and CXCR4) on the transition of chronic inflammation in the upper gastrointestinal tract to neoplasia. CXCR2 ligands, including GRO-α,β,γ/CXCL1,2,3, ENA-78/CXCL5 and IL-8/CXCL8 chemoattract pro-tumoral neutrophils. In addition, angiogenic CXCR2 ligands stimulate the formation of new blood vessels, facilitating tumor progression. The CXCR4 ligand SDF-1/CXCL12 also promotes tumor development by stimulating angiogenesis and by favoring metastasis of CXCR4-positive tumor cells to distant organs producing SDF-1/CXCL12. Furthermore, these angiogenic chemokines also directly enhance tumor cell survival and proliferation. In contrast, the CXCR3 ligands Mig/CXCL9, IP-10/CXCL10 and I-TAC/CXCL11 are angiostatic and attract anti-tumoral T lymphocytes and may therefore mediate tumor growth retardation and regression. Thus, chemokines exert diverging, sometimes dual roles in tumor biology as described for esophageal and gastric cancer. Therefore extensive research is needed to completely unravel the complex chemokine code in specific cancers. Possibly, chemokine-targeted cancer therapy will have to be adapted to the individual's chemokine profile.     

The role of CXC chemokines in the transition of chronic inflammation to esophageal and gastric cancer

Chronic inflammation may increase the risk to develop cancer, for instance esophagitis or gastritis may lead to development of esophageal or gastric cancer, respectively. The key molecules attracting leukocytes to local inflammatory sites are chemokines. We here provide a systematic review on the impact of CXC chemokines (binding the receptors CXCR1, CXCR2, CXCR3 and CXCR4) on the transition of chronic inflammation in the upper gastrointestinal tract to neoplasia. CXCR2 ligands, including GRO-α,β,γ/CXCL1,2,3, ENA-78/CXCL5 and IL-8/CXCL8 chemoattract pro-tumoral neutrophils. In addition, angiogenic CXCR2 ligands stimulate the formation of new blood vessels, facilitating tumor progression. The CXCR4 ligand SDF-1/CXCL12 also promotes tumor development by stimulating angiogenesis and by favoring metastasis of CXCR4-positive tumor cells to distant organs producing SDF-1/CXCL12. Furthermore, these angiogenic chemokines also directly enhance tumor cell survival and proliferation. In contrast, the CXCR3 ligands Mig/CXCL9, IP-10/CXCL10 and I-TAC/CXCL11 are angiostatic and attract anti-tumoral T lymphocytes and may therefore mediate tumor growth retardation and regression. Thus, chemokines exert diverging, sometimes dual roles in tumor biology as described for esophageal and gastric cancer. Therefore extensive research is needed to completely unravel the complex chemokine code in specific cancers. Possibly, chemokine-targeted cancer therapy will have to be adapted to the individual's chemokine profile.     

Transferrin receptor-1 iron-acquisition pathway — Synthesis,kinetics, thermodynamics and rapid cellular internalization of a holotransferrin–maghemite nanoparticle construct

Background

Targeting nanoobjects via the iron-acquisition pathway is always reported slower than the transferrin/receptor endocytosis. Is there a remedy?

Methods

Maghemite superparamagnetic and theragnostic nanoparticles (diameter 8.6 nm) were synthesized, coated with 3-aminopropyltriethoxysilane (NP) and coupled to four holotransferrin (TFe₂) by amide bonds (TFe₂–NP). The constructs were characterized by X-ray diffraction, transmission electron microscopy, FTIR, X-ray Electron Spectroscopy, Inductively Coupled Plasma with Atomic Emission Spectrometry. The in-vitro protein/protein interaction of TFe₂–NP with transferrin receptor-1 (R1) and endocytosis in HeLa cells were investigated spectrophotometrically, by fast T-jump kinetics and confocal microscopy.

Results

In-vitro, R1 interacts with TFe₂–NP with an overall dissociation constant K_D = 11 nM. This interaction occurs in two steps: in the first, the C-lobe of the TFe₂–NP interacts with R1 in 50 μs: second-order rate constant, k₁ = 6 × 10¹⁰ M^− 1 s^− 1; first-order rate constant, k_− 1 = 9 × 10⁴ s^− 1; dissociation constant, K_1d = 1.5 μM. In the second step, the protein/protein adduct undergoes a slow (10,000 s) change in conformation to reach equilibrium. This mechanism is identical to that occurring with the free TFe₂. In HeLa cells, TFe₂–NP is internalized in the cytosol in less than 15 min.

Conclusion

This is the first time that a nanoparticle–transferrin construct is shown to interact with R1 and is internalized in time scales similar to those of the free holotransferrin.

General significance

TFe₂–NP behaves as free TFe₂ and constitutes a model for rapidly targeting theragnostic devices via the main iron-acquisition pathway.