Prior Model Assumptions Research Articles

FASTDXL: A Generalized Screen to Trap Disulfide-Stabilized Complexes for Use in Structural Studies

Structural studies of macromolecular complexes have produced extraordinary insights into a wide variety of biological processes. Unfortunately, as structural biologists pursue larger and more challenging assemblies, weakly stable and/or nonspecific interactions can become significant roadblocks to structure determination. We have developed a rapid and effective pool-based screen, termed FASTDXL (focused array screening technique for disulfide X-linking), to produce and identify disulfide-stabilized protein-nucleic acid assemblies. A significant strength of FASTDXL is that it can take advantage of prior structural knowledge about molecular interactions, but does not necessarily rely upon it. A detailed application of the approach to the difficult problem of trapping a bacterial primase-ssDNA complex is described, validating the method as a route toward obtaining diffracting crystals suitable for structure determination.

Dissecting Effects of Complex Mixtures

Epidemiologic studies commonly investigate multiple correlated exposures, which are difficult to analyze appropriately. Hierarchical modeling provides a promising approach for analyzing such data by adding a higher-level structure or prior model for the exposure effects. This prior model can incorporate additional information on similarities among the correlated exposures and can be parametric, semiparametric, or nonparametric. We discuss the implications of applying these models and argue for their expanded use in epidemiology. While a prior model adds assumptions to the conventional (first-stage) model, all statistical methods (including conventional methods) make strong intrinsic assumptions about the processes that generated the data. One should thus balance prior modeling assumptions against assumptions of validity, and use sensitivity analyses to understand their implications. In doing so - and by directly incorporating into our analyses information from other studies or allied fields - we can improve our ability to distinguish true causes of disease from noise and bias.

Detecting stage-wise outliers in hierarchical Bayesian linear models of repeated measures data

We propose numerical and graphical methods for outlier detection in hierarchical Bayes modeling and analyses of repeated measures regression data from multiple subjects; data from a single subject are generically called a “curve”. The first-stage of our model has curve-specific regression coefficients with possibly autoregressive errors of a prespecified order. The first-stage regression vectors for different curves are linked in a second-stage modeling step, possibly involving additional regression variables. Detection of thestage at which the curve appears to be an outlier and themagnitude and specific component of the violation at that stage is accomplished by embedding the null model into a larger parametric model that can accommodate such unusual observations. We give two examples to illustrate the diagnostics, develop a BUGS program to compute them using MCMC techniques, and examine the sensitivity of the conclusions to the prior modeling assumptions.

Tomographic imaging of the shallow crustal structure of the East Pacific Rise at 9°30′N

Compressional wave travel times from a seismic tomography experiment at 9°30′N on the East Pacific Rise are analyzed by a new tomographic method to determine the three‐dimensional seismic velocity structure of the upper 2.5 km of oceanic crust within a 20×18 km2 area centered on the rise axis. The data comprise the travel times and associated uncertainties of 1459 compressional waves that have propagated above the axial magma chamber. A careful analysis of source and receiver parameters, in conjunction with an automated method of picking P wave onsets and assigning uncertainties, constrains the prior uncertainty in the data to 5 to 20 ms. The new tomographic method employs graph theory to estimate ray paths and travel times through strongly heterogeneous and densely parameterized seismic velocity models. The nonlinear inverse method uses a jumping strategy to minimize a functional that includes the penalty function, horizontal and vertical smoothing constraints, and prior model assumptions; all constraints applied to model perturbations are normalized to remove bias. We use the tomographic method to reject the null hypothesis that the axial seismic structure is two‐dimensional. Three‐dimensional models reveal a seismic structure that correlates well with cross‐ and along‐axis variations in seafloor morphology, the location of the axial summit caldera, and the distribution of seafloor hydrothermal activity. The along‐axis segmentation of the seismic structure above the axial magma chamber is consistent with the hypothesis that mantle‐derived melt is preferentially injected midway along a locally linear segment of the rise and that the architecture of the crustal section is characterized by an en echelon series of elongate axial volcanoes approximately 10 km in length. The seismic data are compatible with a 300‐ to 500‐m‐thick thermal anomaly above a midcrustal melt lens; such an interpretation suggests that hydrothermal fluids may not have penetrated this region in the last 103 years. Asymmetries in the seismic structure across the rise support the inferences that the thickness of seismic layer 2 and the average midcrustal temperature increase to the west of the rise axis. These anomalies may be the result of off‐axis magmatism; alternatively, the asymmetric thermal anomaly may be the consequence of differences in the depth extent of hydrothermal cooling.

A generalized framework for high-frequency optimum target detection in a shallow water, random wave scattering environment

It has been shown that for the case of Gaussian reverberation, the likelihood ratio detector can be made robust with respect to inexact knowledge of the acoustic scattering medium by incorporating a probabilistic description of environmental uncertainty and a model for the wave-number spectrum of the bottom-interacting acoustic field [Premus etal., J. Acoust. Soc. Am. (submitted)]. In this work a generalized form of the optimum detector is presented which requires full-field (amplitude and phase) modeling capability for the scattered field and relies on Monte Carlo integration to treat environmental uncertainty. The performance of the generalized detector formulation is compared to that of the original detector formulation, which is a special case based on explicit modeling of the reverberation second-order statistics that is exact for the case of Gaussian reverberation. Tradeoffs involving the prior knowledge, modeling assumptions and computational requirements are discussed. [Work supported by ONR, Ocean Acoustics.]

Polarization studies on solid electrolyte cells with a full automated galvanostatic current interruption technique

A versatile technique has been described which allows for accurate separation of electrode and electrolyte polarization contributions in solid electrolyte cells. No prior model assumptions are made and polarization behaviour can be studied for currents ranging from 1 μA to 500 mA. The data acquisition and all instrument function are controlled by a microcomputer. Measurements on equivalent circuits simulating solid electrolyte cells indicate that the error in separating electrode and electrolyte polarization contribution was less than 2%. Three important aspects of current flow in zirconia-based solid electrolyte cells, studied by this technique, have been described. These are the stability of the electrode material with respect to temperature and current flow, decomposition of the electrolyte and cell heating especially during heavy current passage.