Kevin K. Anderson

A new approach to automated peak detection

Abstract Spectral peak detection algorithms are often difficult to automate because they either rely on somewhat arbitrary rules, or are tuned to specific spectral peak properties. One popular approach detects peaks where signal intensities exceed some threshold. This threshold is typically set arbitrarily above the noise level or manually by the user. Intensity threshold-based methods can be sensitive to baseline variations and signal intensity. Another popular peak detection approach relies on matching the spectral intensities to a reference peak shape. This approach can be very sensitive to baseline changes and deviations from the reference peak shape. Such methods can be significantly challenged by modern analytical instrumentation where the baseline tends to drift, peaks of interest may have a low signal to noise (S/N) ratio, and no well-defined reference peak shape is available. We present a new approach for spectral peak detection that is designed to be generic and easily automated. Employing a histogram-based model for spectral intensity, peaks are detected by comparing the estimated variance of observations (the x -axis of the spectrum) to the expected variance when no peak is present inside some window of interest. We compare an implementation of this approach to two existing peak detection algorithms using a series of simulated spectra.

Point source detection and characterization for vehicle radiation portal monitors

Many international border crossings presently screen cargo for illicit nuclear material using radiation portal monitors (RPMs) that measure the gamma ray and/or neutron flux emitted by vehicles. The fact that many target sources have a point-like geometry can be exploited to detect subthreshold sources and filter out benign sources that frequently possess a distributed geometry. This report describes a two-step process, which has the potential to complement other alarm algorithms, for detecting and characterizing point sources. The first step applies a matched filter whereas step two uses a weighted nonlinear least squares method. In a base-case simulation, matched filtering detected a 250-cps source injected onto a white-noise background at a 95% detection probability and a 0.003 false alarm probability. For the same simulation, the maximum likelihood estimation technique performed well at source strengths of 250 and 400 cps. These simulations provided a best-case feasibility study for this technique, which will be extended to experimental data that possess false point-source signatures resulting from background shielding caused by vehicle design and cargo distribution

Examination of Count-starved Gamma Spectra Using the Method of Spectral Comparison Ratios

We discuss the determination of energy region (bin) boundaries and decision metrics for gamma-ray spectra, acquired using a mid-resolution detector, that are useful for detecting illicit sources at low total counts. The bins are designed to produce the lowest minimum detectable counts using a spectral comparison ratio technique at a given false-positive rate for a specified population of benign-source spectra. Spectra from the benign source population consist of observations taken by a detector on a moving vehicle, as would be obtained during a search for a missing or hidden source. Raw counts in bins are transformed into a vector of background-corrected count differences. Bin boundaries are determined to yield large values of a standardized length of this vector for benign-plus-benchmark sources by applying an optimization technique. The objective function includes penalties for overlap with the spectral features of naturally occurring radioactive materials. We compare estimated minimum detectable count values for such bins applied to depleted uranium and barium-133 sources with those based on gross counting, and we examine the effect of nuisance potassium-, radium- and thorium-dominated sources. Using this methodology, we demonstrate that energy bins may be chosen to be sensitive to special nuclear materials, improving the likelihood of detection in low-count or masked-source searches.

Affinity Purification of DNA and RNA from Environmental Samples with Peptide Nucleic Acid Clamps

ABSTRACT Bispeptide nucleic acids (bis-PNAs; PNA clamps), PNA oligomers, and DNA oligonucleotides were evaluated as affinity purification reagents for subfemtomolar 16S ribosomal DNA (rDNA) and rRNA targets in soil, sediment, and industrial air filter nucleic acid extracts. Under low-salt hybridization conditions (10 mM NaPO4, 5 mM disodium EDTA, and 0.025% sodium dodecyl sulfate [SDS]) a PNA clamp recovered significantly more target DNA than either PNA or DNA oligomers. The efficacy of PNA clamps and oligomers was generally enhanced in the presence of excess nontarget DNA and in a low-salt extraction-hybridization buffer. Under high-salt conditions (200 mM NaPO4, 100 mM disodium EDTA, and 0.5% SDS), however, capture efficiencies with the DNA oligomer were significantly greater than with the PNA clamp and PNA oligomer. Recovery and detection efficiencies for target DNA concentrations of ≥100 pg were generally >20% but depended upon the specific probe, solution background, and salt condition. The DNA probe had a lower absolute detection limit of 100 fg of target (830 zM [1 zM = 10−21 M]) in high-salt buffer. In the absence of exogenous DNA (e.g., soil background), neither the bis-PNA nor the PNA oligomer achieved the same absolute detection limit even under a more favorable low-salt hybridization condition. In the presence of a soil background, however, both PNA probes provided more sensitive absolute purification and detection (830 zM) than the DNA oligomer. In varied environmental samples, the rank order for capture probe performance in high-salt buffer was DNA > PNA > clamp. Recovery of 16S rRNA from environmental samples mirrored quantitative results for DNA target recovery, with the DNA oligomer generating more positive results than either the bis-PNA or PNA oligomer, but PNA probes provided a greater incidence of detection from environmental samples that also contained a higher concentration of nontarget DNA and RNA. Significant interactions between probe type and environmental sample indicate that the most efficacious capture system depends upon the particular sample type (and background nucleic acid concentration), target (DNA or RNA), and detection objective.

Mixed-effects statistical model for comparative LC-MS proteomics studies.

Comparing a proteins concentrations across two or more treatments is the focus of many proteomics studies. A frequent source of measurements for these comparisons is a mass spectrometry (MS) analysis of a proteins peptide ions separated by liquid chromatography (LC) following its enzymatic digestion. Alas, LC-MS identification and quantification of equimolar peptides can vary significantly due to their unequal digestion, separation, and ionization. This unequal measurability of peptides, the largest source of LC-MS nuisance variation, stymies confident comparison of a proteins concentration across treatments. Our objective is to introduce a mixed-effects statistical model for comparative LC-MS proteomics studies. We describe LC-MS peptide abundance with a linear model featuring pivotal terms that account for unequal peptide LC-MS measurability. We advance fitting this model to an often incomplete LC-MS data set with REstricted Maximum Likelihood (REML) estimation, producing estimates of model goodness-of-fit, treatment effects, standard errors, confidence intervals, and protein relative concentrations. We illustrate the model with an experiment featuring a known dilution series of a filamentous ascomycete fungus Trichoderma reesei protein mixture. For 781 of the 1546 T. reesei proteins with sufficient data coverage, the fitted mixed-effects models capably described the LC-MS measurements. The LC-MS measurability terms effectively accounted for this major source of uncertainty. Ninety percent of the relative concentration estimates were within 0.5-fold of the true relative concentrations. Akin to the common ratio method, this model also produced biased estimates, albeit less biased. Bias decreased significantly, both absolutely and relative to the ratio method, as the number of observed peptides per protein increased. Mixed-effects statistical modeling offers a flexible, well-established methodology for comparative proteomics studies integrating common experimental designs with LC-MS sample processing plans. It favorably accounts for the unequal LC-MS measurability of peptides and produces informative quantitative comparisons of a proteins concentration across treatments with objective measures of uncertainties.

Evaluating concentration estimation errors in ELISA microarray experiments.

BackgroundEnzyme-linked immunosorbent assay (ELISA) is a standard immunoassay to estimate a proteins concentration in a sample. Deploying ELISA in a microarray format permits simultaneous estimation of the concentrations of numerous proteins in a small sample. These estimates, however, are uncertain due to processing error and biological variability. Evaluating estimation error is critical to interpreting biological significance and improving the ELISA microarray process. Estimation error evaluation must be automated to realize a reliable high-throughput ELISA microarray system.In this paper, we present a statistical method based on propagation of error to evaluate concentration estimation errors in the ELISA microarray process. Although propagation of error is central to this method and the focus of this paper, it is most effective only when comparable data are available. Therefore, we briefly discuss the roles of experimental design, data screening, normalization, and statistical diagnostics when evaluating ELISA microarray concentration estimation errors.ResultsWe use an ELISA microarray investigation of breast cancer biomarkers to illustrate the evaluation of concentration estimation errors. The illustration begins with a description of the design and resulting data, followed by a brief discussion of data screening and normalization. In our illustration, we fit a standard curve to the screened and normalized data, review the modeling diagnostics, and apply propagation of error.We summarize the results with a simple, three-panel diagnostic visualization featuring a scatterplot of the standard data with logistic standard curve and 95% confidence intervals, an annotated histogram of sample measurements, and a plot of the 95% concentration coefficient of variation, or relative error, as a function of concentration.ConclusionsThis statistical method should be of value in the rapid evaluation and quality control of high-throughput ELISA microarray analyses. Applying propagation of error to a variety of ELISA microarray concentration estimation models is straightforward. Displaying the results in the three-panel layout succinctly summarizes both the standard and sample data while providing an informative critique of applicability of the fitted model, the uncertainty in concentration estimates, and the quality of both the experiment and the ELISA microarray process.

ProMAT: protein microarray analysis tool

SUMMARY ProMAT is a software tool for statistically analyzing data from enzyme-linked immunosorbent assay microarray experiments. The software estimates standard curves, sample protein concentrations and their uncertainties for multiple assays. ProMAT generates a set of comprehensive figures for assessing results and diagnosing process quality. The tool is available for Windows or Mac, and is distributed as open-source Java and R code. AVAILABILITY ProMAT is available at http://www.pnl.gov/statistics/ProMAT. ProMAT requires Java version 1.5.0 and R version 1.9.1 (or more recent versions). ProMAT requires either Windows XP or Mac OS 10.4 or newer versions.

Calibrating multi-machine power system parameters with the extended Kalman filter

Large-scale renewable resources and novel smart-grid technologies continue to increase the complexity of power systems. As power systems continue to become more complex, accurate modeling for planning and operation becomes a necessity. Inaccurate system models would result in an unreliable assessment of system security conditions and could cause large-scale blackouts. This motivates the need for model parameter calibration, since some or all of the model parameters could either be unknown or inaccurate. In this paper, the extended Kalman filter is used to calibrate the parameters of a multi-machine power system in the presence of faults. The calibration performance is tested under varying fault locations, parameter errors, and measurement noise giving an insight into how many generators and which generators could be difficult to calibrate.

Distributed dynamic state estimation with extended Kalman filter

Increasing complexity associated with large-scale renewable resources and novel smart-grid technologies necessitates real-time monitoring and control. Our previous work applied the extended Kalman filter (EKF) with the use of phasor measurement data (PMU) for dynamic state estimation. However, high computation complexity creates significant challenges for real-time applications. In this paper, the problem of distributed dynamic state estimation is investigated. One domain decomposition method is proposed to utilize decentralized computing resources. The performance of distributed dynamic state estimation is tested on a 16-machine, 68-bus test system.

Time-Spectral Analysis Methods for Spent Fuel Assay Using Lead Slowing-Down Spectroscopy

Nondestructive techniques for measuring the mass of fissile isotopes in spent nuclear fuel is a considerable challenge in the safeguarding of nuclear fuel cycles. A nondestructive assay technology that could provide direct measurement of fissile mass, particularly for the plutonium (Pu) isotopes, and improve upon the uncertainty of todays confirmatory methods is needed. Lead slowing-down spectroscopy (LSDS) has been studied for the spent fuel application previously, but the nonlinear effects of assembly self shielding (of the interrogating neutron population) have led to discouraging assay accuracy for realistic pressurized water reactor fuels. In this paper, we describe the development of time-spectral analysis algorithms for LSDS intended to overcome these self-shielding effects. The algorithm incorporates the tabulated energy-dependent cross sections from key fissile and absorbing isotopes, but leaves their mass as free variables. Multi-parameter regression analysis is then used to directly calculate not only the mass of fissile isotopes in the fuel assembly (e.g., Pu-239, U-235, and Pu-241), but also the mass of key absorbing isotopes such as Pu-240 and U-238. Modeling-based assay results using this self-shielding relationship indicate that LSDS has the potential to directly measure fissile isotopes with less than 5% average relative error for pressurized water reactor assemblies with burnup as high as 60 GWd/MTU. Shortcomings in the initial self-shielding model and potential improvements to the formulation are described.