Semester

Summer

Date of Graduation

2012

Document Type

Thesis

Degree Type

MS

College

Statler College of Engineering and Mineral Resources

Department

Lane Department of Computer Science and Electrical Engineering

Committee Chair

Donald Adjeroh

Abstract

A novel panel of single nucleotide polymorphisms (SNPs) used in the classification of ancestral origin is proposed. The panel is motivated by recent results in human identification and categorization using soft biometrics. The panel makes use of SNPs related to the height and ear size of an individual. The classifier proposed in this thesis is a standard maximum-likelihood classifier that matches an individual to the ancestral profile closest to the individual. This method produces results with lower error rates than other proposed methods.;This thesis proposes a new distinct panel of 55 single nucleotide polymorphisms for determining ancestral origin. This panel was later reduced to two panels of 18SNPs and 22SNPs respectively. This new approach is producing a high success rate, and has great promise for future progress. This panel is unique in that all prior research has made use of SNPs related to skin pigmentation since this is considered the most readily identifiable difference between different ancestral groups. We chose to use the four common ancestry groups of African American, Asian, European, and Sub-Saharan African.;The proposed panel of 55 SNPs had a successful classification rate of 98% when tested on simulated data, while the 22 SNP panel when tested on real data had a successful classification rate of 86% due to SNP availability issues in the real data. Using a knockout procedure, the 55 SNP panel was reduced to an 18 SNP panel that had a successful classification rate of 99.8% on simulated data.

Download

Share

COinS