Lars J. Kangas

Proteomic characterization of nipple aspirate fluid: identification of potential biomarkers of breast cancer.

Mammary ductal cells are the origin for 70–80% of breast cancers. Nipple aspirate fluid (NAF) contains proteins directly secreted by the ductal and lobular epithelium in non-lactating women. Proteomic approaches offer a largely unbiased way to evaluate NAF as a source of biomarkers and are sufficiently sensitive for analysis of small NAF volumes (10–50 µl). In this study, we initially evaluated a new process for obtaining NAF and discovered that this process resulted in a volume of NAF that was suitable for analysis in ∼90% of subjects. Proteomic characterization of NAF identified 64 proteins. Although this list primarily includes abundant and moderately abundant NAF proteins, very few of these proteins have previously been reported in NAF. At least 15 of the NAF proteins identified have previously been reported to be altered in serum or tumor tissue from women with breast cancer, including cathepsin D and osteopontin. In summary, this study provides the first characterization of the NAF proteome and identifies several candidate proteins for future studies on breast cancer markers in NAF.

Three neural network based, sensor systems for environmental monitoring

Compact, portable systems capable of quickly identifying contaminants in the field are of great importance when monitoring the environment. One of the missions of the Pacific Northwest Laboratory is to examine and develop new technologies for environmental restoration and waste management at the Hanford Site (a former Plutonium production facility). In this paper, three prototype sensing systems are discussed. These prototypes are composed of sensing elements, data acquisition system, computer, and neural network implemented in software and are capable of automatically identifying contaminants. The first system employs an array of tin-oxide gas sensors and is used to identify chemical vapors. The second system employs an array of optical sensors and is used to identify the composition of chemical dyes in liquids. The third system contains a portable gamma-ray spectrometer and is used to identify radioactive isotopes. In these systems, the neural network is used to identify the composition of the sensed contaminant. With a neural network, the intense computation takes place during the training process. Once the network is trained, operation consists of propagating the data through the network. Since the computation involved during operation consists of vector-matrix multiplication and application of look-up tables (activation function), unknown samples can be rapidly identified in the field.<<ETX>>

In silico identification software (ISIS)

MOTIVATION Liquid chromatography-mass spectrometry-based metabolomics has gained importance in the life sciences, yet it is not supported by software tools for high throughput identification of metabolites based on their fragmentation spectra. An algorithm (ISIS: in silico identification software) and its implementation are presented and show great promise in generating in silico spectra of lipids for the purpose of structural identification. Instead of using chemical reaction rate equations or rules-based fragmentation libraries, the algorithm uses machine learning to find accurate bond cleavage rates in a mass spectrometer employing collision-induced dissociation tandem mass spectrometry. RESULTS A preliminary test of the algorithm with 45 lipids from a subset of lipid classes shows both high sensitivity and specificity.

Peptide Orientation Affects Selectivity in Ion-Exchange Chromatography

Here we demonstrate that separation of proteolytic peptides, having the same net charge and one basic residue, is affected by their specific orientation toward the stationary phase in ion-exchange chromatography. In electrostatic repulsion−hydrophilic interaction chromatography (ERLIC) with an anion-exchange material, the C-terminus of the peptides is, on average, oriented toward the stationary phase. In cation exchange, the average peptide orientation is the opposite. Data with synthetic peptides, serving as orientation probes, indicate that in tryptic/Lys-C peptides the C-terminal carboxyl group appears to be in a zwitterionic bond with the side chain of the C-terminal Lys/Arg residue. In effect, the side chain is then less basic than the N-terminus, accounting for the specific orientation of tryptic and Lys-C peptides. Analyses of larger sets of peptides, generated from lysates by either Lys-N, Lys-C, or trypsin, reveal that specific peptide orientation affects the ability of charged side chains, such as phosphate residues, to influence retention. Phosphorylated residues that are remote in the sequence from the binding site affect retention less than those that are closer. When a peptide contains multiple charged sites, then orientation is observed to be less rigid and retention tends to be governed by the peptide’s net charge rather than its sequence. These general observations could be of value in confirming a peptide’s identification and, in particular, phosphosite assignments in proteomics analyses. More generally, orientation accounts for the ability of chromatography to separate peptides of the same composition but different sequence.

Machine learning based prediction for peptide drift times in ion mobility spectrometry

MOTIVATION Ion mobility spectrometry (IMS) has gained significant traction over the past few years for rapid, high-resolution separations of analytes based upon gas-phase ion structure, with significant potential impacts in the field of proteomic analysis. IMS coupled with mass spectrometry (MS) affords multiple improvements over traditional proteomics techniques, such as in the elucidation of secondary structure information, identification of post-translational modifications, as well as higher identification rates with reduced experiment times. The high throughput nature of this technique benefits from accurate calculation of cross sections, mobilities and associated drift times of peptides, thereby enhancing downstream data analysis. Here, we present a model that uses physicochemical properties of peptides to accurately predict a peptides drift time directly from its amino acid sequence. This model is used in conjunction with two mathematical techniques, a partial least squares regression and a support vector regression setting. RESULTS When tested on an experimentally created high confidence database of 8675 peptide sequences with measured drift times, both techniques statistically significantly outperform the intrinsic size parameters-based calculations, the currently held practice in the field, on all charge states (+2, +3 and +4). AVAILABILITY The software executable, imPredict, is available for download from http:/omics.pnl.gov/software/imPredict.php CONTACT rds@pnl.gov SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

Identifying at-risk employees: A behavioral model for predicting potential insider threats

A psychosocial model was developed to assess an employee’s behavior associated with an increased risk of insider abuse. The model is based on case studies and research literature on factors/correlates associated with precursor behavioral manifestations of individuals committing insider crimes. In many of these crimes, managers and other coworkers observed that the offenders had exhibited signs of stress, disgruntlement, or other issues, but no alarms were raised. Barriers to using such psychosocial indicators include the inability to recognize the signs and the failure to record the behaviors so that they could be assessed by a person experienced in psychosocial evaluations. We have developed a model using a Bayesian belief network with the help of human resources staff, experienced in evaluating behaviors in staff. We conducted an experiment to assess its agreement with human resources and management professionals, with positive results. If implemented in an operational setting, the model would be part of a set of management tools for employee assessment that can raise an alarm about employees who pose higher insider threat risks. In separate work, we combine this psychosocial model’s assessment with computer workstation behavior to raise the efficacy of recognizing an insider crime in the making.

Fault Detection in Distributed Climate Sensor Networks Using Dynamic Bayesian Networks

The Atmospheric Radiation Measurement (ARM) program operated by the U.S. Department of Energy is one of the largest climate research programs dedicated to the collection of long-term continuous measurements of cloud properties and other key components of the earth’s climate system. Given the critical role that collected ARM data plays in the analysis of atmospheric processes and conditions and in the enhancement and evaluation of global climate models, the production and distribution of high-quality data is one of ARM’s primary mission objectives. Fault detection in ARM’s distributed sensor network is one critical ingredient towards maintaining high quality and useful data. We are modeling ARM’s distributed sensor network as a dynamic Bayesian network where key measurements are mapped to Bayesian network variables. We then define the conditional dependencies between variables by discovering highly correlated variable pairs from historical data. The resultant dynamic Bayesian network provides an automated approach to identifying whether certain sensors are malfunctioning or failing in the distributed sensor network. A potential fault or failure is detected when an observed measurement is not consistent with its expected measurement and the observed measurements of other related sensors in the Bayesian network. We present some of our experiences and promising results with the fault detection dynamic Bayesian network.

Computer-aided tracking and characterization of homicides and sexual assaults (CATCH)

When a serial offender strikes, it usually means that the investigation is unprecedented for that police agency. The volume of incoming leads and pieces of information in the case(s) can be overwhelming as evidenced by the thousands of leads gathered in the Ted Bundy Murders, Atlanta Child Murders, and the Green River Murders. Serial cases can be long term investigations in which the suspect remains unknown and continues to perpetrate crimes. With state and local murder investigative systems beginning to crop up, it will become important to manage that information in a timely and efficient way by developing computer programs to assist in that task. One vital function will be to compare violent crime cases from different jurisdictions so investigators can approach the investigation knowing that similar cases exist. CATCH (Computer Aided Tracking and Characterization of Homicides) is being developed to assist crime investigations by assessing likely characteristics of unknown offenders, by relating a specific crime case to other cases, and by providing a tool for clustering similar cases that may be attributed to the same offenders. CATCH is a collection of tools that assist the crime analyst in the investigation process by providing advanced data mining and visualization capabilities.These tools include clustering maps, query tools, geographic maps, timelines, etc. Each tool is designed to give the crime analyst a different view of the case data. The clustering tools in CATCH are based on artificial neural networks (ANNs). The ANNs learn to cluster similar cases from approximately 5000 murders and 3000 sexual assaults residing in a database. The clustering algorithm is applied to parameters describing modus operandi (MO), signature characteristics of the offenders, and other parameters describing the victim and offender. The proximity of cases within a two-dimensional representation of the clusters allows the analyst to identify similar or serial murders and sexual assaults.

Adaptive life simulator: a novel approach to modeling the cardiovascular system

The adaptive life simulator (ALS) models a subset of the dynamics of the cardiovascular behavior of an individual by using a recurrent artificial neural network. These models are developed for use in applications that require simulations of cardiovascular systems, such as medical mannequins, and in medical diagnostic systems. This approach is unique in that each cardiovascular model is developed from physiological measurements of an individual. Any differences between the modeled variables and the actual variables of an individual can subsequently be used for diagnosis. This approach also exploits sensor fusion applied to biomedical sensors. Sensor fusion optimizes the utilization of the sensors. The advantage of sensor fusion has been demonstrated in applications including control and diagnostics of mechanical and chemical processes.

Charge Prediction of Lipid Fragments in Mass Spectrometry

An artificial neural network is developed for predicting which fragment is charged and which fragment is neutral for lipid fragment pairs produced from a liquid chromatography tandem mass spectrometry simulation process. This charge predictor is integrated into software developed at PNNL for in silico spectra generation and identification of metabolites known as Met ISIS. To test the effect of including charge prediction in Met ISIS, 46 lipids are used which show a reduction in false positive identifications when the charge predictor is utilized.