The evolution of cranial form and function in theropod dinosaurs: insights from geometric morphometrics

Theropod dinosaurs, an iconic clade of fossil species including Tyrannosaurus and Velociraptor, developed a great diversity of body size, skull form and feeding habits over their 160+ million year evolutionary history. Here, we utilize geometric morphometrics to study broad patterns in theropod skull shape variation and compare the distribution of taxa in cranial morphospace (form) to both phylogeny and quantitative metrics of biting behaviour (function). We find that theropod skulls primarily differ in relative anteroposterior length and snout depth and to a lesser extent in orbit size and depth of the cheek region, and oviraptorosaurs deviate most strongly from the "typical" and ancestral theropod morphologies. Noncarnivorous taxa generally fall out in distinct regions of morphospace and exhibit greater overall disparity than carnivorous taxa, whereas large-bodied carnivores independently converge on the same region of morphospace. The distribution of taxa in morphospace is strongly correlated with phylogeny but only weakly correlated with functional biting behaviour. These results imply that phylogeny, not biting function, was the major determinant of theropod skull shape.     

Bayesian survival analysis using a MARS model

A Bayesian multivariate adaptive regression spline fitting approach is used to model univariate and multivariate survival data with censoring. The possible models contain the proportional hazards model as a subclass and automatically detect departures from this. A reversible jump Markov chain Monte Carlo algorithm is described to obtain the estimate of the hazard function as well as the survival curve.     

Multiple spatial-scale resource selection function models in relation to human disturbance for moose in northeastern China

The moose (Alces alces cameloides) population in northeastern China is on the southernmost edge of its distribution in Asia. A survey was conducted to determine moose resource selection and the effects of human disturbance on moose in a study area of 20,661 ha located on the northwestern slope of the Lesser Khingan Mountains, located in northeastern China. Predictive models of resource selection were developed using logistic and autologistic regression. All models considered resource variable selection at two spatial scales, patch and landscape. At the patch scale, moose preferred larger birch (Betula platyphylla) patches, but avoided larger tamarack (Larix gmelinii) patches. At the landscape scale, moose preferred higher densities of tamarack patches, i.e., heterogeneity of tamarack stands, selected areas with more abundant annual shoots, terrain conducive to better concealment, higher altitudes and areas saturated with soil moisture. Roads and forest harvest intervals were identified as important human disturbance factors. This is the first time that moose have been reported to avoid roads, and the avoidance distance was nearly 3 km. We believe that in this region moose under the influence of roads are behaviorally plastic, compared with the indifference of moose to the presence of roads in other regions. Moose avoided forest areas logged more than 3 years previously and preferred areas logged 1–2 years previously. In addition, it may be necessary to monitor the effect of the dynamic of density of roe deer on the spatial distribution of the moose population.     

The log multinomial regression model for nominal outcomes with more than two attributes

An estimate of the risk or prevalence ratio, adjusted for confounders, can be obtained from a log binomial model (binomial errors, log link) fitted to binary outcome data. We propose a modification of the log binomial model to obtain relative risk estimates for nominal outcomes with more than two attributes (the "log multinomial model"). Extensive data simulations were undertaken to compare the performance of the log multinomial model with that of an expanded data multinomial logistic regression method based on the approach proposed by Schouten et al. (1993) for binary data, and with that of separate fits of a Poisson regression model based on the approach proposed by Zou (2004) and Carter, Lipsitz and Tilley (2005) for binary data. Log multinomial regression resulted in "inadmissable" solutions (out-of-bounds probabilities) exceeding 50% in some data settings. Coefficient estimates by the alternative methods produced out-of-bounds probabilities for the log multinomial model in up to 27% of samples to which a log multinomial model had been successfully fitted. The log multinomial coefficient estimates generally had lesser relative bias and mean squared error than the alternative methods. The practical utility of the log multinomial regression model was demonstrated with a real data example. The log multinomial model offers a practical solution to the problem of obtaining adjusted estimates of the risk ratio in the multinomial setting, but must be used with some care and attention to detail.     

Eribulin regresses a doxorubicin‐resistant Ewing's sarcoma with a FUS‐ERG fusion and CDKN2A‐deletion in a patient‐derived orthotopic xenograft (PDOX) nude mouse model

Ewing's sarcoma is a recalcitrant tumor greatly in need of more effective therapy. The aim of this study was to determine the efficacy of eribulin on a doxorubicin (DOX)‐resistant Ewing's sarcoma patient derived orthotopic xenograft (PDOX) model. The Ewing's sarcoma PDOX model was previously established in the right chest wall of nude mice from tumor resected form the patient's right chest wall. In the previous study, the Ewing's sarcoma PDOX was resistant to doxorubicin (DOX) and sensitive to palbociclib and linsitinib. In the present study, the PDOX models were randomized into three groups when the tumor volume reached 60 mm³: G1, untreated control (n = 6); G2, DOX treated (n = 6), intraperitoneal (i.p.) injection, weekly, for 2 weeks); G3, Eribulin treated (n = 6, intravenous (i.v.) injection, weekly for 2 weeks). All mice were sacrificed on day 15. Changes in body weight and tumor volume were assessed two times per week. Tumor weight was measured after sacrifice. DOX did not suppress tumor growth compared to the control group (P = 0.589), consistent with the previous results in the patient and PDOX. Eribulin regressed tumor size significantly compared to G1 and G2 (P = 0.006, P = 0.017) respectively. No significant difference was observed in body weight among any group. Our results demonstrate that eribulin is a promising novel therapeutic agent for Ewing's sarcoma.     

Determination of endogenous faecal phosphorus loss in goats

Four black Liuyang wether goats were fed with corn stover and concentrate formulated to contain four levels of dietary phosphorus (P), including 0.129, 0.140, 0.162 and 0.180% of P. In a 4 × 4 Latin square experiment the endogenous faecal P loss was determined by the regression technique and the substitution method. Treatment effects on faecal and urinary P output, apparent P digestibility and P retention, and saliva P secretion were not significant. A linear relationship was observed between apparent faecal digestible P (Y, g/kg DMI) and P intake (X, g/kg DMI), which was described by the equation: Y = 0.4799X ?0.9209, r² = 0.9869, (p < 0.05). The true P digestibility determined by the regression technique and the substitution method amounted to 48.0 and 48.9%, respectively; the recorded endogenous faecal P losses were 0.92 and 0.93 g/kg DMI, respectively. The study demonstrated the potential of the regression method as well as the substitution method for estimation of true P digestibility and endogenous faecal P losses in goats.