Subset threshold autoregression

We develop in this paper an efficient way to select the best subset threshold autoregressive model. The proposed method uses a stochastic search idea. Differing from most conventional approaches, our method does not require us to fix the delay or the threshold parameters in advance. By adopting the Markov chain Monte Carlo techniques, we can identify the best subset model from a very large of number of possible models, and at the same time estimate the unknown parameters. A simulation experiment shows that the method is very effective. In its application to the US unemployment rate, the stochastic search method successfully selects lag one as the time delay and five best models from more than 4000 choices. Copyright © 2003 John Wiley & Sons, Ltd.     

Sensitivity to antitubulin chemotherapeutics is regulated by MCL1 and FBW7

Microtubules have pivotal roles in fundamental cellular processes and are targets of antitubulin chemotherapeutics. Microtubule-targeted agents such as Taxol and vincristine are prescribed widely for various malignancies, including ovarian and breast adenocarcinomas, non-small-cell lung cancer, leukaemias and lymphomas. These agents arrest cells in mitosis and subsequently induce cell death through poorly defined mechanisms. The strategies that resistant tumour cells use to evade death induced by antitubulin agents are also unclear. Here we show that the pro-survival protein MCL1 (ref. 3) is a crucial regulator of apoptosis triggered by antitubulin chemotherapeutics. During mitotic arrest, MCL1 protein levels decline markedly, through a post-translational mechanism, potentiating cell death. Phosphorylation of MCL1 directs its interaction with the tumour-suppressor protein FBW7, which is the substrate-binding component of a ubiquitin ligase complex. The polyubiquitylation of MCL1 then targets it for proteasomal degradation. The degradation of MCL1 was blocked in patient-derived tumour cells that lacked FBW7 or had loss-of-function mutations in FBW7, conferring resistance to antitubulin agents and promoting chemotherapeutic-induced polyploidy. Additionally, primary tumour samples were enriched for FBW7 inactivation and elevated MCL1 levels, underscoring the prominent roles of these proteins in oncogenesis. Our findings suggest that profiling the FBW7 and MCL1 status of tumours, in terms of protein levels, messenger RNA levels and genetic status, could be useful to predict the response of patients to antitubulin chemotherapeutics.     

A genome-wide association study for celiac disease identifies risk variants in the region harboring IL2 and IL21

We tested 310,605 SNPs for association in 778 individuals with celiac disease and 1,422 controls. Outside the HLA region, the most significant finding (rs13119723; P = 2.0 x 10(-7)) was in the KIAA1109-TENR-IL2-IL21 linkage disequilibrium block. We independently confirmed association in two further collections (strongest association at rs6822844, 24 kb 5' of IL21; meta-analysis P = 1.3 x 10(-14), odds ratio = 0.63), suggesting that genetic variation in this region predisposes to celiac disease.     

Priority and privilege in scientific discovery

The priority rule in science has been interpreted as a behavior regulator for the scientific community, which benefits society by adequately structuring the distribution of intellectual labor across pre-existing research programs. Further, it has been lauded as an intuitively fair way to reward scientists for their contributions, as a special case of society’s “grand reward scheme”. However, we will argue that the current formal framework utilized to model the priority rule idealizes away important aspects of credit attribution, and does so in a way that impacts the conclusions drawn regarding its function in scientific communities. In particular, we consider the social dynamics of credit attribution in order to show that the priority rule can foster structural disadvantages in socially diverse science, as well as drive the distribution of intellectual labor away from optimal.     

Subjective intervention in formal models

Success in forecasting using mathematical/statistical models requires that the models be open to intervention by the user. In practice, a model is only one component of a forecasting system, which also includes the users/forecasters as integral components. Interaction between the user and the model is necessary to adequately cater for events and changes that go beyond the existing form of the model. In this paper we consider Bayesian forecasting models open to interventions, of essentially any form, to incorporate subjective information made available to the user. We discuss principles of intervention and derive theoretical results that provide the means to formally incorporate feedforward interventions into Bayesian models. Two example time series are considered to illustrate why and when such interventions may be necessary to sustain predictive performance.