Resultants in Genetic Linkage Analysis

Ingileif B. Hallgrimsdottir,Bernd Sturmfels

2004-11-18

Abstract: Statistical models for genetic linkage analysis of k-locus diseases are k-dimensional subvarieties of a (3^k-1)-dimensional probability simplex. We determine the algebraic invariants of these models with general characteristics for k=1, in particular we recover, and generalize, the Hardy-Weinberg curve. For k = 2, the invariants are presented as determinants of 32x32-matrices of linear forms in 9 unknowns, a suitable format for computations with numerical data.

Quantitative Methods,Statistics Theory

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to explore the relationship between the shear parameter \( S^* \) and the Reynolds number \( Re \), especially the behavior in turbulent shear flow in a statistically stable state. Specifically, the research aims to answer the following questions:

1. **Relationship between shear parameter and Reynolds number**: Under different experimental conditions, how does the shear parameter \( S^* \) change with the change of the Reynolds number \( Re \)? In particular, when the Reynolds number increases, will the shear parameter tend to saturate?
2. **Decay behavior of high - order derivative moments**: Simulation results in different paths show that the odd - order moments of the transverse derivative (such as skewness) have different decay behaviors with respect to the Reynolds number. Can this difference be explained as different manifestations of local isotropic regression?

### Research Background

- **Homogeneous shear turbulence**: This is a flow model that extends homogeneous, locally isotropic turbulence and is often used in the research between the idealized homogeneous isotropic turbulence of bridges and more realistic channel flows.
- **Influence of shear rate**: There is a constant average shear rate \( S \), which is the source of large - scale anisotropy and has an impact on the statistical properties of small - scale turbulent flow.
- **Experiments and numerical simulations**: Previous studies have shown that at high Reynolds numbers, the decay index of the odd - order normalized transverse derivative moments is greater than - 1, and the co - spectrum of shear stress deviates from the classical - 7/3 power law.

### Main Findings

- **Saturation of shear parameter**: The research shows that for larger Reynolds numbers, the shear parameter \( S^* \) tends to saturate, which is consistent with the prediction based on dimensional analysis.
- **Decay of odd - order moments**: The skewness (the third - order moment) of the transverse derivative has different decay behaviors in different paths. Along path I, the skewness drops rapidly with the Reynolds number; while on path II, it only starts to decline slowly at the maximum Reynolds number or shear parameter value.

### Mathematical Expressions

The relationship between the shear parameter \( S^* \) and the Reynolds number \( Re \) can be expressed by the following formula:
\[ S^*=\frac{S v_{rms}^2}{\epsilon}, \]
where \( S \) is the shear rate, \( v_{rms} \) is the root - mean - square velocity, and \( \epsilon \) is the energy dissipation rate.

The Reynolds number \( Re \) is defined as:
\[ Re = \frac{v_{rms}L_y}{\nu}, \]
where \( L_y \) is the width of the simulation domain and \( \nu \) is the kinematic viscosity.

Through these studies, the author hopes to better understand the regression of local isotropy in shear flow and provide a theoretical basis for future experiments and numerical simulations.