Skip to content

Latest commit

 

History

History
139 lines (106 loc) · 7.1 KB

NEWS.md

File metadata and controls

139 lines (106 loc) · 7.1 KB
Raw

updog 2.1.5

  • Used updated Rcpp to generate RcppExports to fix a CRAN warning.

updog 2.1.4

  • Removed {ggthemes} dependency.
  • Removed usage of ggplot2::aes_string(), since it is deprecated, and replaced it with tidy evaluation idioms.

updog 2.1.3

  • Bug fix: Use && instead of & in C++.

updog 2.1.2

  • Fixed a bug in my use of assertthat::are_equal() and testthat::expect_equal(). See 21 Jan 2022 R-devel/NEWS where it states:

    all.equal.numeric() gains a sanity check on its tolerance argument - calling all.equal(a, b, c) for three numeric vectors is a surprisingly common error.

updog 2.1.1

  1. Added an upper bound to the sequencing error rate in flexdog_full() (and, hence, flexdog() and multidog()). This protects against some poor behavior observed in a corner case. Specifically, F1 populations where the offspring are all the same genotype and is sequenced at moderate to low depth.
  2. Fixed some stale URLs, fixed some style issues found by lintr.

updog 2.1.0

  1. The parallel backend in multidog() is now handled by the future package. If you use the nc argument in multidog(), it should still run in parallel using multiple R sessions on your local machine. However, you can now use the functionality of future to choose your own evaluation strategy, after setting nc = NA. This will also allow you to use schedulers in high performance computing environments through the future.batchtools package. See the multidog() function documentation for more details.
  2. The vignette "Genotyping Many SNPs with multidog()" goes through an example of using the future package.
  3. A new experimental function, export_vcf(), is in the works to export multidog objects to a VCF file. This is not yet exported because I still have a few bugs to fix.
  4. plot.multidog() will now plot the parent read-counts in F1 and S1 populations.
  5. Internally, multidog() now uses iterators through the iterators package to send only subsets of the data to each R process.
  6. A new internal .combine function is used in the foreach() call of multidog() in order to decrease the memory usage of multidog().

updog 2.0.2

This is a massive edit of the updog software. Major changes include:

  1. No more support of model = "ash". It seemed that model = "norm" was always better and faster, so I just got rid of the "ash" option. This also extremely simplified the code.
  2. Removal of mupdog(). I think this was a good idea, but the computation was way too slow to be usable.
  3. Revision of model = "f1pp" and model = "s1pp". These now include interpretable parameterizations that are meant to be identified via another R package. But support is only for tetraploids right now.
  4. multidog() now prints some nice ASCII art when it's run.
  5. format_multidog() now allows you to format multiple variables in terms of a multidimensional array.
  6. Fixes a bug where format_multidog() was reordering the SNP dimensions. This was fine as long as folks used dimnames properly, but now it should allow folks to also use dim positions.
  7. Updog now returns genotype log-likelihoods.

updog 1.2.1

  • Adds filter_snp() for filtering the output of multidog() based on predicates in terms of the variables in snpdf.
  • Removes stringr from Imports. I was only using it in one place so I replaced that code with base R code.
  • Removes Rmpfr from Suggests. No longer needed since CVXR is no longer suggested.

updog 1.2.0

  • Adds multidog() for genotyping multiple SNPs using parallel computing.
  • Adds plot.multidog() for plotting the output of multidog().
  • Adds format_multidog() for formatting the output of multidog() to be a matrix.
  • Removes dependency on CVXR. This makes install and maintenance a little easier. The defaults for this specific problem were a little faster anyway.
  • No longer changes the color scale in plot_geno() based on what genotypes are present.
  • In .cpp files, we now coerce objects to be unsigned before comparing. This gets rid of some warnings during install.

updog 1.1.3

  • Updates documentation to include the Bioinformatics publication, Gerard and Ferrão (2020) <doi:10.1093/bioinformatics/btz852>.
  • Adds the "internal" keyword to functions that most users don't need.
  • Removes the tidyverse from the Suggests field. I was only using this in the vignettes, so I changed it to base R (except for ggplot2).

updog 1.1.1

  • Updates documentation to include Gerard and Ferrão (2020) <doi:10.1101/751784> as a reference.
  • Minor fixes to documentation.

updog 1.1.0

  • Introduces more flexible priors for more general populations.
  • Places a normal prior distribution on the logit of the overdispersion parameter. This might change genotype calls from previous versions of updog. To reproduce the genotype calls from previous versions of updog, simply set mean_od = 0 and var_od = Inf in flexdog().
  • Adds the method = "custom" option to flexdog(). This lets users choose the genotype distribution if it is completely known a priori.
  • Documentation updates.

updog 1.0.1

  • Fixes a bug with option model = "s1pp" in flexdog(). I was originally not constraining the levels of preferential pairing to be the same in both segregations of the same parent. This is now fixed. But the downside is that model = "s1pp" is now only supported for ploidy = 4 or ploidy = 6. This is because the optimization becomes more difficult for larger ploidy levels.

  • I fixed some documentation. Perhaps the biggest error comes from this snippet from the original documentation of flexdog:

    The value of prop_mis is a very intuitive measure for the quality of the SNP. prop_mis is the posterior proportion of individuals mis-genotyped. So if you want only SNPS that accurately genotype, say, 95% of the individuals, you could discard all SNPs with a prop_mis under 0.95.

    This now says

    The value of prop_mis is a very intuitive measure for the quality of the SNP. prop_mis is the posterior proportion of individuals mis-genotyped. So if you want only SNPS that accurately genotype, say, 95% of the individuals, you could discard all SNPs with a prop_mis over 0.05.

  • I've now exported some C++ functions that I think are useful. You can call them in the usual way.

updog 0.99.0

  • This is a complete re-working of the code in updog. The old version may be found in the updogAlpha package.
  • The main function is now flexdog().
  • An experimental approach mupdog() is now live. We provide no guarantees about mupdog()'s performance.
  • Oracle misclassification error rates may be calculated in oracle_mis().
  • Genotypes can be simulated using rgeno().
  • Next-generation sequencing data can be simulated using rflexdog().