A Swift, Versatile Phylogenomic and High-Throughput Sequencing Simulator.

jackalope

Overview

For studies using high-throughput sequencing (HTS) data, simulations can be vital for planning sampling design and testing bioinformatic tools. However, most HTS sequencing tools provide only very simple ways of adding deviations from a reference genome. For HTS studies that focus on patterns of genomic variation among individuals, populations, or species, having a tool that can simulate realistic patterns of molecular evolution and generate HTS data from those simulations would be quite useful.

jackalope simply and efficiently simulates (i) haplotypes from reference genomes and (ii) reads from both Illumina and Pacific Biosciences (PacBio) platforms. It can either read reference genomes from FASTA files or simulate new ones. Variant haplotypes can be simulated using summary statistics, phylogenies, Variant Call Format (VCF) files, and coalescent simulations—the latter of which can include selection, recombination, and demographic fluctuations. jackalope can simulate single, paired-end, or mate-pair Illumina reads, as well as reads from Pacific Biosciences These simulations include sequencing errors, mapping qualities, multiplexing, and optical/PCR duplicates. All outputs can be written to standard file formats.

Installation

Dependencies

Before installing jackalope, you should update the packages Rhtslib and zlibbioc. Since both of these are on Bioconductor, you should update BiocManager, too.

if (!requireNamespace("BiocManager", quietly = TRUE) ||
    "BiocManager" %in% row.names(old.packages())) {
  install.packages("BiocManager")
}
BiocManager::install(c("Rhtslib", "zlibbioc"))

Stable version

# To install the latest stable version from CRAN:
install.packages("jackalope")

Development version

# install.packages("devtools")
remotes::install_github("lucasnell/jackalope")

Enabling OpenMP

To use multithreading in jackalope, you’ll need to compile it from source using the proper flags. If you’ve enabled OpenMP properly, running jackalope:::using_openmp() in R should return TRUE.

Windows and Linux

The first step is to add the following to the .R/Makevars (.R/Makevars.win on Windows) file inside the home directory:

PKG_CXXFLAGS += $(SHLIB_OPENMP_CXXFLAGS)
PKG_CFLAGS += $(SHLIB_OPENMP_CFLAGS)
PKG_LIBS += $(SHLIB_OPENMP_CFLAGS)

Then, you should be able to install jackalope by running the following in R:

install.packages("jackalope", type = "source")
## Or, for development version:
# remotes::install_github("lucasnell/jackalope")

macOS, R version >= 4.0.0

Follow the directions here to install R compiler tools: https://thecoatlessprofessor.com/programming/cpp/r-compiler-tools-for-rcpp-on-macos/.

Check your version of gcc using gcc --version in the Terminal. Then, check the table at https://mac.r-project.org/openmp/ to see which version of the runtime OpenMP downloads you need. For LLVM version 9.0.1, you run the following in the Terminal:

export version="9.0.1"

curl -O https://mac.r-project.org/openmp/openmp-${version}-darwin17-Release.tar.gz
sudo tar fvx openmp-${version}-darwin17-Release.tar.gz -C /

For the next step of actually installing jackalope, one option is to add the following to your ~/.R/Makevars file:

CPPFLAGS += -Xclang -fopenmp
LDFLAGS += -lomp

… then installing jackalope by running install.packages("jackalope", type = "source") or remotes::install_github("lucasnell/jackalope") in R.

This might not be desirable since it affects all package installations. An alternative method is to use the package withr:

withr::with_makevars(c(CPPFLAGS = "-Xclang -fopenmp", LDFLAGS = "-lomp"), 
                     install.packages("jackalope", type = "source"),
                     ## For development version:
                     # remotes::install_github("lucasnell/jackalope"),
                     assignment = "+=")

macOS, R version >= 3.4* and < 4.0.0

Add the following to the .R/Makevars file inside the home directory:

PKG_CXXFLAGS += $(SHLIB_OPENMP_CXXFLAGS)
PKG_CFLAGS += $(SHLIB_OPENMP_CFLAGS)
PKG_LIBS += $(SHLIB_OPENMP_CFLAGS)

Next, go to https://cran.r-project.org/bin/macosx/tools/ and download the newest versions of (1) the clang compiler (version 8 at the time of writing) and (2) GNU Fortran (version 6.1 at the time of writing). The downloads will have the .pkg extension. Next, install clang and gfortran by opening these .pkg files and following the directions.

After this, add the following to your ~/.R/Makevars file (replacing clang8 with your version of the clang compiler):

CLANG8=/usr/local/clang8/bin/clang
CC=$(CLANG8)
CXX=$(CLANG8)++
CXX11=$(CLANG8)++
CXX14=$(CLANG8)++
CXX17=$(CLANG8)++
CXX1X=$(CLANG8)++
LDFLAGS=-L/usr/local/clang8/lib

Now you should be able to install jackalope by running install.packages("jackalope", type = "source") in R.

For more information, please see https://thecoatlessprofessor.com/programming/openmp-in-r-on-os-x/.

Usage

Below shows how to simulate a 10kb genome, then create haplotypes from that genome using a phylogenetic tree:

library(jackalope)
reference <- create_genome(n_chroms = 10, len_mean = 1000)
tr <- ape::rcoal(5)
ref_haplotypes <- create_haplotypes(reference, haps_phylo(tr), sub_JC69(0.1))
ref_haplotypes
#>                               << haplotypes object >>
#> # Haplotypes: 5
#> # Mutations: 16,870
#> 
#>                           << Reference genome info: >>
#> < Set of 10 chromosomes >
#> # Total size: 10,000 bp
#>   name                             chromosome                             length
#> chrom0     CTGGCATTGAATCATATGAGGTGGCCAT...ACGTTGCACGATTGATTAAATTCCTGAA      1000
#> chrom1     CACTCCGTCGCACACTAGGTTTCGAGAT...GTGAGCTCGCGTACATGGAGCATTCTGT      1000
#> chrom2     CTTAGCCGGAGCGACTCGGAGCAACTGC...TGGCGTAATATGCCAGGTCCCGCGTGGC      1000
#> chrom3     CGCCTTCCATTTAGGACTTGTATTGGTG...GCTAAACTCCATGTGACTGTAATGTCAG      1000
#> chrom4     GGGTGATATGGTGTGCATGCTGAATTCG...AGAGTCTAGAGTCTCTGGGAGGTCAGGT      1000
#> chrom5     TTCGTTGGTGGGTGTCCTATGCTACGAT...CGCCCGCCGGTTTGACTTACTCGATTGG      1000
#> chrom6     GCATGGACAGATGTGATCTGAGTATACG...CAGACCCCATAAGGCCTGGGACACTGTG      1000
#> chrom7     TCGTTTCAACGTCCTTAAGTGTAGTATC...GGCTCGTTAGCTCTCCGAGGAGACGAGG      1000
#> chrom8     CAGGTAAGTTATCAAAGAACCTTCCTGG...ACGCATCACCTCGCAAGGAGACTCGTTA      1000
#> chrom9     GGTAGTAATTAGGCTTAAAATAGCAGTG...ATAACAAATGTTCGGCATACGATCTACG      1000

Below simulates 500 million paired-end, 100 bp reads from the haplotypes:

illumina(ref_haplotypes, out_prefix = "illumina", n_reads = 500e6,
         paired = TRUE, read_length = 100)

Below simulates 500 thousand PacBio reads from the reference genome:

pacbio(ref, out_prefix = "pacbio", n_reads = 500e3)