GoldPolish (aka GoldRush-Edit)

GoldPolish (aka GoldRush-Edit) is an efficient draft genome assembly polishing tool that uses long reads for polishing. ntEdit polishes the draft assembly and flags additional erroneous regions, then Sealer fills assembly gaps and erroneous sequence regions flagged by ntEdit. The polisher is adapted from the ntedit_sealer_protocol to use long reads instead of short reads.

Dependencies

• Build

  • GCC 7+ or Clang 8+ (with OpenMP support)
  • meson
  • ninja
  • btllib v1.6.2+
  • boost

• Run

  • GNU Make
  • Python 3
  • btllib v1.6.2+
  • ntLink v1.3.5+
  • minimap2
  • snakemake
  • intervaltree

The dependencies can be installed through Conda package manager:

conda install -c conda-forge -c bioconda compilers meson ninja boost-cpp btllib ntlink minimap2 snakemake intervaltree

Citation

If you use GoldPolish in your research, please cite:

Wong J, Coombe L, Nikolić V, Zhang E, Nip KM, Sidhu P, Warren RL and Birol I (2023). Linear time complexity de novo long read genome assembly with GoldRush. Nature Communications, 14(1), 2906. https://doi.org/10.1038/s41467-023-38716-x

Installation

To build GoldPolish and install it at $GOLDPOLISH_PREFIX, run the following commands from within the goldpolish directory:

meson setup build --buildtype release --prefix $GOLDPOLISH_PREFIX
cd build
ninja install

Usage

To polish a draft assembly named assembly.fa with long reads named reads.fa and store the results at assembly-polished.fa, run the following:

goldpolish assembly.fa reads.fa assembly-polished.fa

You can run goldpolish --help to see the available options:

usage: goldpolish [-h] [-k K] [-b BSIZE] [-m SHARED_MEM] [-t THREADS] [-v] [-x MX_MAX_READS_PER_10KBP] [-s SUBSAMPLE_MAX_READS_PER_10KBP]
                  [--ntlink | --minimap2 | --mappings MAPPINGS]
                  seqs_to_polish polishing_seqs output_seqs

positional arguments:
  seqs_to_polish        Sequences to polish.
  polishing_seqs        Sequences to polish with.
  output_seqs           Filename to write polished sequences to.

optional arguments:
  -h, --help            show this help message and exit
  -k K                  k-mer sizes to use for polishing. Example: -k32 -k28 (Default: 32, 28, 24, 20)
  -b BSIZE, --bsize BSIZE
                        Batch size. A batch is how many polished sequences are processed per Bloom filter. (Default: 1)
  -m SHARED_MEM, --shared-mem SHARED_MEM
                        Shared memory path to do polishing in. (Default: /dev/shm)
  -t THREADS, --threads THREADS
                        How many threads to use. (Default: 48)
  -v, --verbose
  -x MX_MAX_READS_PER_10KBP, --mx-max-reads-per-10kbp MX_MAX_READS_PER_10KBP
                        When subsampling, increase the common minimizer count threshold for ntLink mappings until there's at most this many reads per 10kbp of polished sequence.
                        (Default: 150)
  -s SUBSAMPLE_MAX_READS_PER_10KBP, --subsample-max-reads-per-10kbp SUBSAMPLE_MAX_READS_PER_10KBP
                        Random subsampling of mapped reads. For ntLink mappings, this is done after common minimizer subsampling. For minimap2 mappings, only this subsampling is done.
                        By default, 40 if using minimap2 mappings and 100 if using ntLink mappings.
  --ntlink              Run ntLink to generate read mappings (default).
  --minimap2            Run minimap2 to generate read mappings.
  --mappings MAPPINGS   Use provided pre-generated mappings. Accepted formats are PAF, SAM, and *.verbose_mapping.tsv from ntLink.
  --target              Run GoldPolish in targeted mode
  -l LENGTH, --length LENGTH
                        GoldPolish-Target flank length (if --target specified) (Default: 64)
  --bed BED             BED file specifying target coordinates (if --target specified)
  --softmask            Target coordinates determined from softmasked regions in the input assembly (if --target specified)
  --k-ntlink            k-mer size used for ntLink mappings (if --ntlink specified) (Default: 88)
  --w-ntlink            Window size used for ntLink mappings (if --ntlink specified) (Default: 1000)

GoldPolish-Target

GoldPolish can be run in a targeted mode, polishing specified regions of the assembly (either by looking for softmasked sequences or by using target coordinates that are specified in a BED file). To run GoldPolish-Target, use the --target flag for the GoldPolish command.

To polish a draft assembly named assembly.fa with long reads named reads.fa, where targeted polishing coordinates are stored in a BED file named polishing_coordinates.bed, and store the results in a FASTA file called assembly-polished.fa, run the following:

goldpolish --target --bed polishing_coordinates.bed assembly.fa reads.fa assembly-polished.fa