EVOLUTION-MANAGER
Edit File: fitstahl.html
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><title>R: Fit Stahl interference model</title> <meta http-equiv="Content-Type" content="text/html; charset=utf-8" /> <link rel="stylesheet" type="text/css" href="R.css" /> </head><body> <table width="100%" summary="page for fitstahl {qtl}"><tr><td>fitstahl {qtl}</td><td style="text-align: right;">R Documentation</td></tr></table> <h2>Fit Stahl interference model</h2> <h3>Description</h3> <p>Fit the Stahl model for crossover inference (or the chi-square model, which is a special case). </p> <h3>Usage</h3> <pre> fitstahl(cross, chr, m, p, error.prob=0.0001, maxit=4000, tol=1e-4, maxm=15, verbose=TRUE) </pre> <h3>Arguments</h3> <table summary="R argblock"> <tr valign="top"><td><code>cross</code></td> <td> <p>An object of class <code>cross</code>. See <code><a href="read.cross.html">read.cross</a></code> for details.</p> </td></tr> <tr valign="top"><td><code>chr</code></td> <td> <p>Optional vector indicating the chromosomes to consider. This should be a vector of character strings referring to chromosomes by name; numeric values are converted to strings. Refer to chromosomes with a preceding <code>-</code> to have all chromosomes but those considered. A logical (TRUE/FALSE) vector may also be used.</p> </td></tr> <tr valign="top"><td><code>m</code></td> <td> <p>Interference parameter (a non-negative integer); if unspecified, this is estimated.</p> </td></tr> <tr valign="top"><td><code>p</code></td> <td> <p>The proportion of chiasmata coming from the no interference mechanism in the Stahl model (0 <= p <= 1). p=0 gives the chi-square model. If unspecified, this is estimated.</p> </td></tr> <tr valign="top"><td><code>error.prob</code></td> <td> <p>The genotyping error probability. If = NULL, it is estimated.</p> </td></tr> <tr valign="top"><td><code>maxit</code></td> <td> <p>Maximum number of iterations to perform.</p> </td></tr> <tr valign="top"><td><code>tol</code></td> <td> <p>Tolerance for determining convergence.</p> </td></tr> <tr valign="top"><td><code>maxm</code></td> <td> <p>Maximum value of m to consider, if m is unspecified.</p> </td></tr> <tr valign="top"><td><code>verbose</code></td> <td> <p>Logical; indicates whether to print tracing information.</p> </td></tr> </table> <h3>Details</h3> <p>This function is currently only available for backcrosses and intercrosses. </p> <p>The Stahl model of crossover interference (of which the chi-square model is a special case) is fit. In the chi-square model, points are tossed down onto the four-strand bundle according to a Poisson process, and every <i>(m+1)</i>st point is a chiasma. With the assumption of no chromatid interference, crossover locations on a random meiotic product are obtained by thinning the chiasma process. The parameter <i>m</i> (a non-negative integer) governs the strength of crossover interference, with <i>m=0</i> corresponding to no interference. </p> <p>In the Stahl model, chiasmata on the four-strand bundle are a superposition of chiasmata from two mechanisms, one following a chi-square model and one exhibiting no interference. An additional parameter, <i>p</i>, gives the proportion of chiasmata from the no interference mechanism. </p> <p>If all of <code>m</code>, <code>p</code>, and <code>error.prob</code> are specified, any of them with length > 1 must all have the same length. </p> <p>If <code>m</code> is unspecified, we do a grid search starting at 0 and stop when the likelihood decreases (thus assuming a single mode), or <code>maxm</code> is reached. </p> <h3>Value</h3> <p>A matrix with four columns: m, p, error.prob, and the log likelihood. </p> <p>If specific values for m, p, error.prob are provided, the log likelihood for each set are given. </p> <p>If some are left unspecified, the maximum likelihood estimates are provided in the results. </p> <h3>Author(s)</h3> <p>Karl W Broman, <a href="mailto:broman@wisc.edu">broman@wisc.edu</a> </p> <h3>References</h3> <p>Armstrong, N. J., McPeek, M. J. and Speed, T. P. (2006) Incorporating interference into linkage analysis for experimental crosses. <em>Biostatistics</em> <b>7</b>, 374–386. </p> <p>Zhao, H., Speed, T. P. and McPeek, M. S. (1995) Statistical analysis of crossover interference using the chi-square model. <em>Genetics</em> <b>139</b>, 1045–1056. </p> <h3>See Also</h3> <p><code><a href="est.map.html">est.map</a></code>, <code><a href="sim.cross.html">sim.cross</a></code></p> <h3>Examples</h3> <pre> # Simulate genetic map: one chromosome of length 200 cM with # a 2 cM marker spacing mymap <- sim.map(200, 51, anchor.tel=TRUE, include.x=FALSE, sex.sp=FALSE, eq.spacing=TRUE) # Simulate data under the chi-square model, no errors mydata <- sim.cross(mymap, n.ind=250, type="bc", error.prob=0, m=3, p=0) # Fit the chi-square model for specified m's ## Not run: output <- fitstahl(mydata, m=1:5, p=0, error.prob=0) plot(output$m, output$loglik, lwd=2, type="b") # Find the MLE of m in the chi-square model ## Not run: mle <- fitstahl(mydata, p=0, error.prob=0) ## Not run: # Simulate data under the Stahl model, no errors mydata <- sim.cross(mymap, n.ind=250, type="bc", error.prob=0, m=3, p=0.1) # Find MLE of m for the Stahl model with known p mle.stahl <- fitstahl(mydata, p=0.1, error.prob=0) # Fit the Stahl model with unknown p and m, # get results for m=0, 1, 2, ..., 8 output <- fitstahl(mydata, m=0:8, error.prob=0) plot(output$m, output$loglik, type="b", lwd=2) ## End(Not run) </pre> <hr /><div style="text-align: center;">[Package <em>qtl</em> version 1.66 <a href="00Index.html">Index</a>]</div> </body></html>