Main Page
Deanship
The Dean
Dean's Word
Curriculum Vitae
Contact the Dean
Vision and Mission
Organizational Structure
Vice- Deanship
Vice- Dean
KAU Graduate Studies
Research Services & Courses
Research Services Unit
Important Research for Society
Deanship's Services
FAQs
Research
Staff Directory
Files
Favorite Websites
Deanship Access Map
Graduate Studies Awards
Deanship's Staff
Staff Directory
Files
Researches
Contact us
عربي
English
About
Admission
Academic
Research and Innovations
University Life
E-Services
Search
Deanship of Graduate Studies
Document Details
Document Type
:
Thesis
Document Title
:
PARALLEL ANALYSIS OF DNA PROFILE MIXTURES WITH A LARGE NUMBER OF CONTRIBUTORS
تنفيذ موازٍ لتحليل تنميط الحمض النووي مكون من خليط كبير من المساهمين
Subject
:
Faculty of Computing and Information Technology
Document Language
:
Arabic
Abstract
:
Determining the number of contributors in a DNA profile is a popular practice in various forensic laboratories. For example, it helps in cases of sexual assault when the source of DNA mixture can combine different individuals such as the victim, the criminal, and the victim partner. This problem (also known as DNA profiling) is one of the hardest in the forensic science domain. The complexity would increase as the number of unknown contributors would increase. A few methods and their software implementations have been developed to address this problem. However, its (exponential) computational complexity has been the major deterring factor holding its advancements and applications. Some work on OpenMP based parallel implementation of DNA profiling exists but it has only been applied to small problems and is not scalable. The aim of this thesis is to improve the performance of DNA profiling tools using parallel computing, enabling the interpretation of mixtures with a larger number of unknowns within a shorter time frame. We developed two different implementations of DNA profiling focusing on the maximum likelihood computations. The first implementation is based on the OpenMP paradigm. We were able to achieve faster multicore implementations by a factor of up to three compared to the best software tool available (i.e. NOCIt) globally. The second implementation is based on the hybrid OpenMP/MPI implementation. This is the first ever hybrid implementation of DNA profiling enabling the likelihood ratio computations to scale to virtually any number of unknowns. We report results of the hybrid implementation for up to 10 unknowns delivering over 52x performance over OpenMP, using up to 3072 cores, reducing the likelihood computation time from 5.8 days to 2.7 hours. In the coming years, the complete genome sequencing technologies are expected to be available in a single or only a few cells, and this is likely to change the DNA profiling landscape.
Supervisor
:
Dr. Ayad Ahmed Mohammed Saleh Al-Bishri
Thesis Type
:
Master Thesis
Publishing Year
:
1439 AH
2018 AD
Co-Supervisor
:
Dr. Rashed Mahmoud
Added Date
:
Tuesday, February 6, 2018
Researchers
Researcher Name (Arabic)
Researcher Name (English)
Researcher Type
Dr Grade
Email
عماد محمد العمودي
Al Amoudi, Emad Mohammed
Researcher
Master
Files
File Name
Type
Description
43056.pdf
pdf
Back To Researches Page