Table of Contents

• Back to Case Studies

• Acknowledgements

• Introduction

• Data Description

  • Data (space delimited)

• Objectives

• FAQs

• References

Blood Pressure

CASE STUDY - FAQs

Last modified 2003-05-20 23:06

Q: What is the definition of hypertension?
A: Hypertension is defined as SPB > 140mmHg.

Q: "Treatment" is Y, N in the data description but 0,1 in the file.
A: The numerical equivalence for Yes / No is 1 / 0.

Q: Are the markers on the same chromosome?

A: Markers next to each other can be considered as coming from the same chromosome and markers far apart 

    can be considered from different chromosomes.

Q: Could you provide us with some more information regarding the independence of the observations from

     these markers?

A: Markers next to each other are more correlated and than markers far away.

Q: How genetic markers are simulated?

A: Samples from natural populations show non-independence between alleles and genotypes at different

     genetic loci measured by "linkage disequilibrium", LD (Crow, Kimura, 1970). This non-independence is a

     consequence of various evolutionary forces, including genetic drift, mutation process, migration, and

     selection. A large population at equilibrium approaches independence (lack of LD). Recombination process
     facilitates this process, therefore loci in physical proximity of each other tend to show larger LD. The extent 

     of LD varies across human populations. There is usually substantial LD at distances over tens of kilobases.

     LD in this data set was modeled by a Markov process that results in LD decay detectable over distances of

     one to three markers.  Marginally, population allele frequencies in this simulation follow U(0,1) distribution

     with LD between neighbouring markers reaching 0.5 to 1 of its maximum possible value. First, population

     frequencies have been created.  Actual samples were obtained by multinomial sampling.