# EstDynamics: Estimate TE dynamics using mismatch data In TE: Insertion/Deletion Dynamics for Transposable Elements

## Description

Given the number of mismatches and element lengths for an LTR retrotransposon family, estimate the age distribution, insertion rate, and deletion rates.

## Usage

 ```1 2 3 4 5 6 7``` ```EstDynamics(mismatch, len, r = 0.013, perturb = 2, rateRange = NULL, plotFit = FALSE, plotSensitivity = FALSE, pause = plotFit && plotSensitivity, main = sprintf("n = %d", n)) EstDynamics2(mismatch, len, r = 0.013, nTrial = 10L, perturb = 2, rateRange = NULL, plotFit = FALSE, plotSensitivity = FALSE, pause = plotFit && plotSensitivity, ...) ```

## Arguments

 `mismatch` A vector containing the number of mismatches. `len` A vector containing the length of each element. `r` Mutation rate (substitutions/(million year * site)) used in the calculation. `perturb` A scalar multiple to perturb the estimated death rate from the null hypothesis estimate. Used to generate the sensitivity analysis. `rateRange` A vector of death rates, an alternative to `perturb` for specifying the death rates. `plotFit` Whether to plot the distribution fits. `plotSensitivity` Whether to plot the sensitivity analysis. `pause` Whether to pause after each plot. `main` The title for the plot. `nTrial` The number of starting points for searching for the MLE. `...` Pass to EstDynamics

## Details

`EstDynamics` estimates the TE dynamics through fitting a negative binomial fit to the mismatch data, while `EstDynamics2` uses a mixture model. For detailed implementation see References.

## Value

`EstDynamics` returns a `TEfit` object, containing the following fields, where the unit for time is million years ago (Mya):

 `pvalue` The p-value for testing H_0: The insertion rate is uniform over time. `ageDist` A list containing the estimated age distributions. `insRt` A list containing the estimated insertion rates. `agePeakLoc` The maximum point (in age) of the age distribution. `insPeakLoc` The maximum point (in time) of the insertion rate. `estimates` The parameter estimates from fitting the distributions; see References `sensitivity` A list containing the results for the sensitivity analysis, with fields `time`: time points; `delRateRange`: A vector for the range of deletion rates; `insRange`: A matrix whose columns contain the insertion rates under different scenarios. `n` The sample size. `meanLen` The mean of element length. `meanDiv` The mean of divergence. `KDE` A list containing the kernel density estimate for the mismatch data. `logLik` The log-likelihoods of the parametric fits.

This function returns a `TEfit2` object, containing all the above fields for `TEfit` and the following:

 `estimates2` The parameter estimates from fitting the mixture distribution. `ageDist2` The estimated age distribution from fitting the mixture distribution. `insRt2` The estimated insertion rate from fitting the mixture distribution. `agePeakLoc2` Maximum point(s) for the age distribution. `insPeakLoc2` Maximum point(s) for the insertion rate.

## References

Dai, X., Wang, H., Dvorak, J., Bennetzen, J., Mueller, H.-G. (2018). "Birth and Death of LTR Retrotransposons in Aegilops tauschii". Genetics

## Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18``` ```# Analyze Gypsy family 24 (Nusif) data(AetLTR) dat <- subset(AetLTR, GroupID == 24 & !is.na(Chr)) set.seed(1) res1 <- EstDynamics(dat\$Mismatch, dat\$UngapedLen, plotFit=TRUE, plotSensitivity=FALSE, pause=FALSE) # p-value for testing a uniform insertion rate res1\$pvalue # Use a mixture distribution to improve fit res2 <- EstDynamics2(dat\$Mismatch, dat\$UngapedLen, plotFit=TRUE) # A larger number of trials is recommended to achieve the global MLE ## Not run: res3 <- EstDynamics2(dat\$Mismatch, dat\$UngapedLen, plotFit=TRUE, nTrial=1000L) ## End(Not run) ```

TE documentation built on May 1, 2019, 10:13 p.m.