No abstract

Background Despite increasing data on the impact of the microbiome on cancer, the dynamics and role of the microbiome in infection during acute myelogenous leukemia (AML) therapy are unknown. Thus, we sought to determine relationships between microbiome composition and infectious outcomes in AML patients receiving induction chemotherapy (IC). Methods Buccal and fecal specimens (478 samples) were collected twice weekly from 34 AML patients undergoing IC. Oral and stool microbiomes were characterized by 16S rRNA V4 sequencing using Illumina MiSeq. Microbial diversity and genera composition were associated with clinical outcomes. Results Baseline stool α-diversity was significantly lower in patients that developed infections during IC compared to those that did not (P = 0.047). Significant decreases in both oral and stool microbial α-diversity were observed over the course of IC, with a linear correlation between α-diversity change at the two sites (P = 0.02). Loss of both oral and stool α-diversity was significantly associated with carbapenem receipt (P < 0.01). Domination events by the majority of genera were transient (median duration = 1 sample), while the number of domination events by pathogenic genera significantly increased over the course of IC (P=0.002). Moreover, patients who lost microbial diversity over the course of induction chemotherapy were significantly more likely to contract a microbiologically documented infection within the 90 days post-IC neutrophil recovery (P=0.04). Conclusion These data present the largest longitudinal analyses of oral and stool microbiomes in AML patients and suggest that microbiome measurements could assist with mitigation of infectious complications of AML therapy.

Summary This article develops a unified theoretical and computational framework for false discovery control in multiple testing of spatial signals. We consider both point-wise and cluster-wise spatial analyses, and derive oracle procedures which optimally control the false discovery rate, false discovery exceedance and false cluster rate, respectively. A data-driven finite approximation strategy is developed to mimic the oracle procedures on a continuous spatial domain. Our multiple testing procedures are asymptotically valid and can be effectively implemented using Bayesian computational algorithms for analysis of large spatial data sets. Numerical results show that the proposed procedures lead to more accurate error control and better power performance than conventional methods. We demonstrate our methods for analyzing the time trends in tropospheric ozone in eastern US.

SUMMARYMany models for the study of point-referenced data explicitly introduce spatial random effects to capture residual spatial association. These spatial effects are customarily modelled as a zeromean stationary Gaussian process. The spatial Dirichlet process introduced by Gelfand et al. (2005) produces a random spatial process which is neither Gaussian nor stationary. Rather, it varies about a process that is assumed to be stationary and Gaussian. The spatial Dirichlet process arises as a probability-weighted collection of random surfaces. This can be limiting for modelling and inferential purposes since it insists that a process realization must be one of these surfaces. We introduce a random distribution for the spatial effects that allows different surface selection at different sites. Moreover, we can specify the model so that the marginal distribution of the effect at each site still comes from a Dirichlet process. The development is offered constructively, providing a multivariate extension of the stick-breaking representation of the weights. We then introduce mixing using this generalized spatial Dirichlet process. We illustrate with a simulated dataset of independent replications and note that we can embed the generalized process within a dynamic model specification to eliminate the independence assumption.

Background: For more than two decades, national career development programs (CDPs) have addressed underrepresentation of women faculty in academic medicine through career and leadership curricula. We evaluated CDP participation impact on retention. Methods: We used Association of American Medical Colleges data to compare 3268 women attending CDPs from 1988 to 2008 with 17,834 women and 40,319 men nonparticipant faculty similar to CDP participants in degree, academic rank, first year of appointment in rank, and home institution. Measuring from first year in rank to departure from last position held or December 2009 (study end date), we used Kaplan-Meier curves; Cox survival analysis adjusted for age, degree, tenure, and department; and 10-year rates to compare retention. Results: CDP participants were significantly less likely to leave academic medicine than their peers for up to 8 years after appointment as Assistant and Associate Professors. Full Professor participants were significantly less likely to leave than non-CDP women. Men left less often than non-CDP women at every rank. Participants attending more than one CDP left less often than those attending one, but results varied by rank. Patterns of switching institutions after 10 years varied by rank; CDP participants switched significantly less often than men at Assistant and Associate Professor levels and significantly less often than non-CDP women among Assistant Professors. Full Professors switched at equal rates. Conclusion: National CDPs appear to offer retention advantage to women faculty, with implications for faculty performance and capacity building within academic medicine. Intervals of retention advantage for CDP participants suggest vulnerable periods for intervention.