JSM 2005 Online Program

Abstract #304493

This is the preliminary program for the 2005 Joint Statistical Meetings in Minneapolis, Minnesota. Currently included in this program is the "technical" program, schedule of invited, topic contributed, regular contributed and poster sessions; Continuing Education courses (August 7-10, 2005); and Committee and Business Meetings. This on-line program will be updated frequently to reflect the most current revisions.

To View the Program:
You may choose to view all activities of the program or just parts of it at any one time. All activities are arranged by date and time.

The views expressed here are those of the individual authors
and not necessarily those of the ASA or its board, officers, or staff.

The Program has labeled the meeting rooms with "letters" preceding the name of the room, designating in which facility the room is located:

Minneapolis Convention Center = “MCC” Hilton Minneapolis Hotel = “H” Hyatt Regency Minneapolis = “HY”

Back to main JSM 2005 Program page

Legend:

= Applied Session,

= Theme Session,

= Presenter

Activity Number:	277
Type:	Contributed
Date/Time:	Tuesday, August 9, 2005 : 10:30 AM to 12:20 PM
Sponsor:	Biometrics Section
Abstract - #304493
Title:	Accelerating the EM Algorithm for Haplotype Reconstruction from Population Genotype Data
Author(s):	Hormuzd Katki*+ and Ravi Varadhan and Christophe Roland and Bingshu E. Chen and Philip S. Rosenberg
Companies:	National Cancer Institute and Johns Hopkins University and Université des Sciences et Technologies de Lille and National Institutes of Health and National Cancer Institute
Address:	6120 Executive Blvd Room 8044 MSC 7244, Rockville, MD, 20852-4910, United States
Keywords:	Haplotype ; Genotype ; convergence acceleration ; EM Algorithm ; extrapolation methods
Abstract:	Because current genotyping methods rarely provide haplotype information, many statistical methods have been proposed to reconstruct haplotypes given genotype information. One popular method derives maximum-likelihood estimates using the EM algorithm. In many realistic cases, this EM algorithm converges slowly. We apply the novel class of SQUAREM schemes for accelerating EM algorithms (Roland and Varadhan 2004). We present simulations that compare the speed and convergence properties of SQUAREM and EM in a variety of realistic haplotype scenarios. SQUAREM provides useful accelerations, on average reducing the number of EM-step evaluations by a factor of five. The overhead cost of implementing SQUAREM schemes is minimal because they require only the updated parameters from the EM-step. Occasionally, SQUAREM (or even EM) will not converge to the global maximum, but this usually can be circumvented by starting the algorithms from different random starting values. These accelerations are handy for computationally-intensive analyses such as simulations, bootstrapping, and permutation testing.

The address information is for the authors that have a + after their name.
Authors who are presenting talks have a * after their name.

Back to the full JSM 2005 program