Development of predictive models of laboratory animal growth using artificial neural networks

Traditional regression analysis of body weight growth curvesencounters problems .when the data are extremely variable. Whiletransformations are often employed to meet the criteria of theanalysis, some transformations are inadequate for normalizingthe data. Regression analysis also requires presuppositionsregarding the model to be fit and the techniques to be usedin the analysis. An alternative approach using artificial neuralnetworks is presented which may be suitable for developing predictivemodels of growth. Neural networks are simulators of the processesthat occur in the biological brain during the learning process.They are trained on the data, developing the necessary algorithmswithin their internal architecture, and produce a predictivemodel based on the learned facts. A dataset of Sprague–Dawleyrat (Rattus norvegicus) weights is analyzed by both traditionalregression analysis and neural network training. Predictionsof body weight are made from both models. While both methodsproduce models that adequately predict the body weights, theneural network model is superior in that it combines accuracyand precision, being less influenced by longitudinal variabilityin the data. Thus, the neural network provides another toolfor researchers to analyze growth curve data.     

Molecular Characterization and Mapping of Murine Genes Encoding Three Members of the Stefin Family of Cysteine Proteinase Inhibitors

Stefins or Type 1 cystatins belong to a large, evolutionarily conserved protein superfamily, the members of which inhibit the papain-like cysteine proteinases. We report here on the molecular cloning and chromosomal localization of three newly identified members of the murine stefin gene family. These genes, designated herein as mouse stefins 1, 2, and 3, were isolated on the basis of their relatively increased expression in moth-eaten viable compared to normal congenic mouse bone marrow cells. The open reading frames of the stefin cDNAs encode proteins of approximately 11.5 kDa that show between 50 and 92% identity to sequences of stefins isolated from various other species. Data from Southern analysis suggest that the murine stefin gene family encompasses at least 6 and possibly 10-20 members, all of which appear to be clustered in the genome. Analysis of interspecific backcross mice indicates that the genes encoding the three mouse stefins all map to mouse chromosome 16, a localization that is consistent with the recent assignment of the human stefin A gene to a region of conserved homology between human chromosome 3q and the proximal region of mouse chromosome 16.     

First chromosome scale genomes of ithomiine butterflies (Nymphalidae: Ithomiini): Comparative models for mimicry genetic studies

The ithomiine butterflies (Nymphalidae: Danainae) represent the largest known radiation of Müllerian mimetic butterflies. They dominate by number the mimetic butterfly communities, which include species such as the iconic neotropical Heliconius genus. Recent studies on the ecology and genetics of speciation in Ithomiini have suggested that sexual pheromones, colour pattern and perhaps hostplant could drive reproductive isolation. However, no reference genome was available for Ithomiini, which has hindered further exploration on the genetic architecture of these candidate traits, and more generally on the genomic patterns of divergence. Here, we generated high-quality, chromosome-scale genome assemblies for two Melinaea species, M. marsaeus and M. menophilus, and a draft genome of the species Ithomia salapia. We obtained genomes with a size ranging from 396 to 503 Mb across the three species and scaffold N50 of 40.5 and 23.2 Mb for the two chromosome-scale assemblies. Using collinearity analyses we identified massive rearrangements between the two closely related Melinaea species. An annotation of transposable elements and gene content was performed, as well as a specialist annotation to target chemosensory genes, which is crucial for host plant detection and mate recognition in mimetic species. A comparative genomic approach revealed independent gene expansions in ithomiines and particularly in gustatory receptor genes. These first three genomes of ithomiine mimetic butterflies constitute a valuable addition and a welcome comparison to existing biological models such as Heliconius, and will enable further understanding of the mechanisms of adaptation in butterflies.     

Seasonal incidence and ecology of the tick Ixodes ricinus (Acari: Ixodidae) on grazing pastures in Western France

A longitudinal survey was carried out during a 2 year period in Western France to assess the infestation level of grazing pastures byIxodes ricinus ticks. Four farms were visited once a month and each of the grazing pastures was sampled in the centre and at the border using the blanket dragging method. A total of 3562I. ricinus (34 adults, 900 nymphs and 2628 larvae) were collected and the infestation was significantly higher during the first year (p<0.0001). The infestation level byI. ricinus varied between grazing pastures and farms. Grazing pastures in the vicinity of forest were more infested than the others, all through the study. The seasonal distribution of ticks showed peaks, with low fluctuations between farms, years and stages. Tick abundance could not be related to vegetation, but only to the vicinity of woods.     

A motile but non-swarming mutant of Proteus mirabilis lacks FlgN, a facilitator of flagella filament assembly

A Tn phoA mutant of Proteus mirabilis was isolated, which had lost the ability to swarm, yet was still motile. The transposon had inserted into flgN , a flagella gene encoding a 147-amino-acid protein of undefined function. Proteus flgN is arranged in an operon with the class III anti-σ²⁸ gene, flgM , flanked by the class II genes, flgA , flgBCD and flhBA , and a novel putative virulence-related gene. The flgN mutation caused a substantial reduction in cell surface-associated flagellin, particularly during differentiation to the normally hyperflagellated swarm cell. This was not due to an effect on flagella gene expression or a typical defect in the flagella export apparatus as there was no class III gene downregulation by FlgM feedback, or intracellular flagellin accumulation. Loss of FlgN nevertheless caused a severe reduction in the incorporation of pulse-labelled flagellin into the membrane/flagellum fraction of differentiating cells. Substantial amounts of both non-oligomeric flagellin and flagellin degradation products appeared in the extracellular medium, although the few mature filaments made by the mutant were no more sensitive to proteolysis than those of the wild type. FlgN appeared soluble and active in the cytosol. The data suggest that the function of FlgN is to facilitate the initiation of flagella filament assembly, a role that may be especially critical in attaining the much higher concentration of surface flagellin required for swarming. Proteus FlgN has leucine zipper-like motifs arranged on potential amphipathic helices, a feature conserved in cytosolic chaperones for the exported substrates of flagella-related type III virulence systems. While gel filtration of FlgN from the soluble cell fraction did not establish an interaction with flagellin, it indicated that FlgN may associate with an unknown component and/or form an oligomer.     

Frameshift deletions of exons 3-7 and revertant fibers in Duchenne muscular dystrophy: mechanisms of dystrophin production.

Duchenne muscular dystrophy (DMD) patients with mutations that disrupt the translational reading frame produce little or no dystrophin. Two exceptions are the deletion of exons 3-7 and the occurrence of rare dystrophin-positive fibers (revertant fibers) in muscle of DMD patients. Antibodies directed against the amino-terminus and the 5' end of exon 8 did not detect dystrophin in muscle from patients who have a deletion of exons 3-7. However, in all cases, dystrophin was detected with an antibody directed against the 3' end of exon 8. The most likely method of dystrophin production in these cases is initiation at a new start codon in exon 8. We also studied two patients who have revertant fibers: one had an inherited duplication of exons 5-7, which, on immunostaining, showed two types of revertant fibers; and the second patient had a 2-bp nonsense mutation in exon 51, which creates a cryptic splice site. An in-frame mRNA that uses this splice site in exon 51 was detected. Immunostaining demonstrated the presence of the 3' end of exon 51, which is in agreement with the use of this mRNA in revertant fibers. The most likely method of dystrophin production in these fibers is a second mutation that restores the reading frame.     

An electrogenic proton pump in plasma membranes from the cellular slime mould Dictyostelium discoideum

Plants and fungi possess an outwardly directed plasma membrane proton pump that may regulate intracellular pH. We provide the first demonstration that amoebae of the slime mould Dictyostelium discoideum also possess a similar proton pump. It can be assayed either as an ATPase activity in highly purified plasma membranes or as a proton pump, after solubilization and reconstruction into liposomes. The pump is inhibited by vanadate, diethylstilbestrol (DES) and miconazole but not by azide or ouabain. The proton pump described here may represent the target for the action of DES and miconazole, both of which have previously been shown to induce stalk cell formation during the in vitro development of Dictyostelium.     

Effects of insulin, somatomedin--a multiplication stimulating activity (MSA)--and transferrin, on the lipolysis of rat adipocytes in vitro

Antilipolytic effect was researched when insulin (0.1 and 1 mIU/ml), MSA (200 and 500 ng/ml) and transferrin (2 and 5 micrograms/ml) were added to a suspension of freshly isolated rat adipocytes in vitro. Lipolysis was measured as glycerol secretion in the medium: micromoles/90 minutes/100 mg total lipids. Insulin (1 mIU/ml) reduced adrenalinic stimulation of lipolysis: A 1 microgram/ml (P less than 0.05). MSA 200 ng/ml had no effect. MSA 500 ng/ml reduced basal lipolysis and adrenalinic stimulation (P less than 0.05), and increased insulin-induced antilipolysis (P less than 0.05). Transferrin was active, only when insulin is present: antilipolysis increased (P less than 0.05).