Umashanger Thayasivam

Potential utility of autoantibodies as blood-based biomarkers for early detection and diagnosis of Parkinson's disease.

INTRODUCTION There is a great need to identify readily accessible, blood-based biomarkers for Parkinsons disease (PD) that are useful for accurate early detection and diagnosis. This advancement would allow early patient treatment and enrollment into clinical trials, both of which would greatly facilitate the development of new therapies for PD. METHODS Sera from a total of 398 subjects, including 103 early-stage PD subjects derived from the Deprenyl and Tocopherol Antioxidative Therapy of Parkinsonism (DATATOP) study, were screened with human protein microarrays containing 9,486 potential antigen targets to identify autoantibodies potentially useful as biomarkers for PD. A panel of selected autoantibodies with a higher prevalence in early-stage PD was identified and tested using Random Forest for its ability to distinguish early-stage PD subjects from controls and from individuals with other neurodegenerative and non-neurodegenerative diseases. RESULTS Results demonstrate that a panel of selected, blood-borne autoantibody biomarkers can distinguish early-stage PD subjects (90% confidence in diagnosis) from age- and sex-matched controls with an overall accuracy of 87.9%, a sensitivity of 94.1% and specificity of 85.5%. These biomarkers were also capable of differentiating patients with early-stage PD from those with more advanced (mild-moderate) PD with an overall accuracy of 97.5%, and could distinguish subjects with early-stage PD from those with other neurological (e.g., Alzheimers disease and multiple sclerosis) and non-neurological (e.g., breast cancer) diseases. CONCLUSION These results demonstrate, for the first time, that a panel of selected autoantibodies may prove to be useful as effective blood-based biomarkers for the diagnosis of early-stage PD.

Detection of Alzheimer's disease at mild cognitive impairment and disease progression using autoantibodies as blood-based biomarkers

There is an urgent need to identify biomarkers that can accurately detect and diagnose Alzheimers disease (AD). Autoantibodies are abundant and ubiquitous in human sera and have been previously demonstrated as disease‐specific biomarkers capable of accurately diagnosing mild‐moderate stages of AD and Parkinsons disease.

Autoantibodies as diagnostic biomarkers for the detection and subtyping of multiple sclerosis

The goal of this preliminary proof-of-concept study was to use human protein microarrays to identify blood-based autoantibody biomarkers capable of diagnosing multiple sclerosis (MS). Using sera from 112 subjects, including 51 MS subjects, autoantibody biomarkers effectively differentiated MS subjects from age- and gender-matched normal and breast cancer controls with 95.0% and 100% overall accuracy, but not from subjects with Parkinsons disease. Autoantibody biomarkers were also useful in distinguishing subjects with the relapsing-remitting form of MS from those with the secondary progressive subtype. These results demonstrate that autoantibodies can be used as noninvasive blood-based biomarkers for the detection and subtyping of MS.

Detection of anomalies in network traffic using L 2 E for accurate speaker recognition

Recently, widespread use of digital speech communication has spawned a multitude of Voice over IP (VoIP) applications. These applications require the ability to identify speakers in real time. One of the challenges in accurate speaker recognition is the inability to detect anomalies in network traffic generated by attacks on VoIP applications. This paper presents L2E, an innovative approach to detect anomalies in network traffic for accurate speaker recognition. The L2E method is capable of online speaker recognition from live packet streams of voice packets by performing fast classification over a defined subset of the features available in each voice packet. The experimental results show that L2E is highly scalable and accurate in detecting a wide range of anomalies in network traffic.

Robust speaker verification with a two classifier format and feature enhancement

In the presence of environmental noise, speaker verification systems inevitably see a decrease in performance. This paper proposes the (1) use of two parallel classifiers, (2) feature enhancement based on blind signal-to-noise ratio (SNR) estimation and (3) fusion, to improve the performance of speaker verification systems. The two classifiers are based on Gaussian mixture models and the partial least-squares technique. Speech corrupted by additive noise at SNRs from 0 to 30 dB are used for authentication. A two-way analysis of variance validates the performance gain offered by the methods used. The outputs of the classifiers are fused together in different ways. The fusion method where the scores of the classifiers are added together is found to be the best method again using statistical analysis.

Accuracy of class prediction using similarity functions in PAM

Clustering have been proven to be an effective technique for finding data instances with similar characteristics. Such algorithms are based on the notion of distance between data points, often computed using Euclidean metric. That is why, clustering algorithms are mostly applicable to the data sets comprising of numerical values. However, the real life data often consist of features which are categorical in nature. For example, to identify abnormal behavior or a cyberattack in a network, we usually examine packet headers which contain categorical values such as source and destination IP addresses, source and destination port numbers, upper layer protocols, etc. Euclidean metric is not applicable to such data sets because it cannot compute the distance between categorical variables. To address this problem, similarity functions have been designed to determine the relationship between given categorical values. Similarity defines how closely related the objects are to one another. Often similarity could be thought of as opposite to distance where similar objects have high value, while dissimilar objects have low or zero value. In this paper we explored accuracy of various similarity functions using the Partitioning Around Medoids (PAM) clustering algorithm. We tested similarity functions on several data sets to determine their ability to correctly predict the class labels. We also examined the applicability of various similarity functions to different types of data sets.

A Comparison of Supervised Learning Techniques for Clustering

The significance of data mining has experienced dramatic growth over the past few years. This growth has been so drastic that many industries and academic disciplines apply data mining in some form. Data mining is a broad subject that encompasses several topics and problems; however this paper will focus on the supervised learning classification problem and discovering ways to optimize the classification process. Four classification techniques (naive Bayes, support vector machine, decision tree, and random forest) were studied and applied to data sets from the UCI Machine Learning Repository. A Classification Learning Toolbox (CLT) was developed using the R statistical programming language to analyze the date sets and report the relationships and prediction accuracy between the four classifiers.

Robust L_{2}E Parameter Estimation of Gaussian Mixture Models: Comparison with Expectation Maximization

The purpose of this paper is to discuss the use of \(L_{2}E\) estimation that minimizes integrated square distance as a practical robust estimation tool for unsupervised clustering. Comparisons to the expectation maximization (EM) algorithm are made. The \(L_{2}E\) approach for mixture models is particularly useful in the study of big data sets and especially those with a consistent numbers of outliers. The focus is on the comparison of \(L_{2}E\) and EM for parameter estimation of Gaussian Mixture Models. Simulation examples show that the \(L_{2}E\) approach is more robust than EM when there is noise in the data (particularly outliers) and for the case when the underlying probability density function of the data does not match a mixture of Gaussians.