Con-Gen 2020. Eric C. Anderson’s resources

Tuesday, September 8, 2020

The presentation follows the narrative in the following html document:

• https://eriqande.github.io/con-gen-2020/bayes-mcmc-gtyperr-narrative.nb.html

Within that notebook are links to the Shiny Apps I have written to play with genotype likelihoods (as well as directions on how to use them). If you want to get the associated R project, you can download it from Box at:

• https://umt.box.com/s/jdsq97eg2tsse3mw4ig5f3h2ynmdhg74

or get it directly from GitHub:

• https://github.com/eriqande/ngs-genotype-models

And, only as a last resort, if you can’t get the Shiny Apps working in RStudio on your computer, you can visit their pages on shinyapps.io:

• https://eriqande.shinyapps.io/ngs-genotype-models/001-allele-freq-estimation.Rmd
• https://eriqande.shinyapps.io/ngs-genotype-models/002-genotype-likelihoods-from-reads.Rmd

Monday, Sept. 14, 2020

Genotyping Errors in RAD data

We will continue with the narrative from the week before, but starting from estimating genotyping errors from departures from Hardy-Weinberg equilibrium:

• https://eriqande.github.io/con-gen-2020/bayes-mcmc-gtyperr-narrative.nb.html#estimating-genotyping-error-rates-in-some-rad-datasets

After that short presentation, we will have a break-out room, hands-on practical session interlude:

Hands-on practical with whoa

1. Log on to the RStudio server.
2. At the R console do:

library(whoa)

• Then, in the RStudio Help panel (lower right), type whoa in the text search box and hit return.
• Click the “Index” link at the bottom of the result.
• The click the “User guides, package vignettes,…” link.
• Choose the whoa tutorial and run through it.
• (Note that this convoluted path to get to the vignette is necessary because of a limitation of RStudio Server…)

Further whoa fun

In the RStudio terminal from your home directory, do this:

cp -r instructor_materials/Eric_Anderson/whoa-practical-session ./

Then open the RStudio project inside there and follow the directions in: 001-whoa-first-steps.Rmd.

Amplicon Sequencing and Microhaplotypes

When you need really accurate genotype calls on many individuals, approaches that yield high read depths at a smaller number of targets are useful. I will talk about our use of microhaplotypes, following these slides:

• https://eriqande.github.io/con-gen-2020/microhaps.pdf

Hands-on practical with microhaplot

This can be run from using the ConGen 2020 remote desktop servers:

• Log in to the remote desktop using Microsoft Remote Desktop.
• Open the Terminal Emulator app (black square app in the dock at bottom of screen).
• Type R into the terminal and hit return (this starts an R session).
• Within R, type these lines:

library(microhaplot)
mvShinyHaplot("~/Shiny")
vignette("microhaplot-walkthrough")

The final line there should open a package vignette in a Firefox tab. This vignette provides instructions for the hands-on session. * Note that the vignette is not well rendered in Firefox on the Remote Desktop. You can get a better looking version straight from CRAN on your own Desktop: https://cran.r-project.org/web/packages/microhaplot/vignettes/microhaplot-walkthrough.html

• Launch the microhaplot Shiny app by issuing this command in R:

runShinyHaplot("~/Shiny/microhaplot")

That will open a FireFox tab with the Shiny App. * From the dropdown box, choose the fish2.rds data set.

Now, you can go through the vignette, starting from “About the Example Data”. This section provides a lot of description about the different panels/views of the microhaplotype data that are available. If you want to just get going and start clicking on stuff, you might try proceeding to the section Mission one: An overview of loci in the vignette.

Getting microhaplot on your own computer

To do this, go to https://github.com/ngthomas/microhaplot and follow the installation instructions.