Context based mixture model for cell phase identification in automated fluorescence microscopy

Background  

Automated identification of cell cycle phases of individual live cells in a large population captured via automated fluorescence microscopy technique is important for cancer drug discovery and cell cycle studies. Time-lapse fluorescence microscopy images provide an important method to study the cell cycle process under different conditions of perturbation. Existing methods are limited in dealing with such time-lapse data sets while manual analysis is not feasible. This paper presents statistical data analysis and statistical pattern recognition to perform this task.     

Bioluminescence imaging reveals inhibition of tumor cell proliferation by Alzheimer's amyloid β protein

Background  

Cancer and Alzheimer's disease (AD) are two seemingly distinct diseases and rarely occur simultaneously in patients. To explore molecular determinants differentiating pathogenic routes towards AD or cancer, we investigate the role of amyloid β protein (Aβ) on multiple tumor cell lines that are stably expressing luciferase (human glioblastoma U87; human breast adenocarcinoma MDA-MB231; and mouse melanoma B16F).     

The automated, accurate and reproducible determination of steady-state permeation parameters from percutaneous permeation data

Procedures for the in vitro determination of percutaneous permeation with Franz diffusion cells are widely accepted. However, the calculation of relevant endpoints, such as the steady-state flux (J) and the permeation coefficient (P(app)), still depends on visual data inspection or an approximation of the steady-state flux as the maximum observed absorption rate. As both these approaches must be considered inappropriate, an automated and reproducible algorithm to analyse permeation data is presented. The method detects both lag-times and non-linear data resulting from substance accumulation in the acceptor compartment of static diffusion cells. It was evaluated by using simulated data, and data from experiments with caffeine and testosterone on bovine udder skin and human reconstituted epidermis (SkinEthic), which represent model barriers with high and low barrier strengths, respectively. It was shown that the algorithm is a suitable method for the identification of steady-state ranges in permeation data. If used on data generated with appropriate experimental approaches, it is a reproducible and time-saving alternative to the visual analysis of diffusion data.     

The use of reconstructed human epidermis for skin absorption testing: Results of the validation study

A formal validation study was performed, in order to investigate whether the commercially-available reconstructed human epidermis (RHE) models, EPISKIN, EpiDerm and SkinEthic, are suitable for in vitro skin absorption testing. The skin types currently recommended in the OECD Test Guideline 428, namely, ex vivo human epidermis and pig skin, were used as references. Based on the promising outcome of the prevalidation study, the panel of test substances was enlarged to nine substances, covering a wider spectrum of physicochemical properties. The substances were tested under both infinite-dose and finite-dose conditions, in ten laboratories, under strictly controlled conditions. The data were subjected to independent statistical analyses. Intra-laboratory and inter-laboratory variability contributed almost equally to the total variability, which was in the same range as that in preceding studies. In general, permeation of the RHE models exceeded that of human epidermis and pig skin (the SkinEthic RHE was found to be the most permeable), yet the ranking of substance permeation through the three tested RHE models and the pig skin reflected the permeation through human epidermis. In addition, both infinite-dose and finite-dose experiments are feasible with RHE models. The RHE models did not show the expected significantly better reproducibility, as compared to excised skin, despite a tendency toward lower variability of the data. Importantly, however, the permeation data showed a sufficient correlation between all the preparations examined. Thus, the RHE models, EPISKIN, EpiDerm and SkinEthic, are appropriate alternatives to human and pig skin, for the in vitro assessment of the permeation and penetration of substances when applied as aqueous solutions.     

Tobamovirus evolution: gene overlaps, recombination, and taxonomic implications

Tobamoviruses, mostly isolated from solanaceous plants, may represent ancient virus lineages that have codiverged with their hosts. Recently completed nucleotide sequences of six nonsolanaceous tobamoviruses allowed assessment of the codivergence hypothesis and support a third subgroup within tobamoviruses. The genomic sequences of 12 tobamoviruses and the partial sequences of 11 others have been analyzed. Comparisons of the predicted protein sequences revealed three clusters of tobamoviruses, corresponding to those infecting solanaceous species (subgroup 1), those infecting cucurbits and legumes (subgroup 2), and those infecting crucifers. The orchid-infecting odontoglossum ringspot tobamovirus was associated with subgroup 1 genomes by its coat and movement protein sequences, but with the crucifer-pathogenic tobamoviruses by the remainder of its genome, suggesting that it is the progeny of a recombinant. For four of five genomic regions, subgroup 1 and 3 genomes were equidistant from a subgroup 2 genome chosen for comparison, suggesting uniform rates of evolution. A phylogenetic tree of plant families based on the tobamoviruses they harbor was congruent with that based on rubisco sequences but had a different root, suggesting that codivergence was tempered by rare events of viruses of one family colonizing another family. The proposed subgroup 3 viruses probably have an origin of virion assembly in the movement protein gene, a large (25-codon) overlap of movement and coat protein open reading frames, and a comparably shorter genome. Codon-position- dependent base compositions and codon prevalences suggested that the coat protein frame of the overlap region was ancestral. Bootstrapped parsimony analysis of the nucleotides in the overlap region and of the sequences translated from the -1 frame (the subgroup 3 movement protein frame) of this region produced trees inconsistent with those deduced from other regions. The results are consistent with a model in which a no or short overlap organization was ancestral. Despite encoding of subgroup 2 and 3 movement protein C-termini by nonhomologous nucleotides, weak similarities between their amino acid sequences suggested convergent sequence evolution.