Powered by

# Simulate Linked Reads

In 

Simulate linked reads from a genome

You may want to benchmark haplotag data on different kinds of genomic variants. To do that, you'll need known variants (like those created by simulate {snpindel,...} ) and linked-read sequences. In Harpy v1.x this was done using a modified version of LRSIM, however, Harpy v2.x now uses the purpose-built software Mimick (originally XENIA from the VISOR project). Mimick does exactly what you would need it to do, so to keep the familiarity of simulate linkedreads in Harpy, we just expose a very thinly veiled wrapper for Mimick. The only additions here are that Harpy automatically installs Mimick and can dispatch the job to an HPC using --hpc like other workflows and ensures proper pairing and haplotype concatenation, otherwise it really just runs Mimick exactly as you would. All of Mimick's command line arguments are exposed to Harpy, except --mutations, -indels, and -extindels, which are set to 0 to make sure you are only simulating linked-reads exactly.

Rather than having to maintain two copies of the same documentation, please head over to the Mimick documentation.

usage
harpy simulate linkedreads OPTIONS... BARCODES FASTA...
example
harpy simulate linkedreads -t 4 18,96 data/genome.hap1.fasta data/genome.hap2.fasta

# Running Options

long name default type description
BARCODES required input barcode file or length,count
FASTA required input fasta file(s)
--output-prefix -o simulated/SIM output file prefix
--output-type -O standard output format of FASTQ files
--quiet -q 0 0 all output, 1 single progress bar, 2 no output
--regions -r one or more regions to simulate, in BED format
--threads -t 2 number of threads to use for simulation
--coverage 30 mean coverage target for simulated data
--distance 500 integer outer distance between the two ends in bp
--error 0.02 0-1 base error rate, fixed at this number
--extindels 0 0-1 hidden indel extension rate
--indels 0 0-1 hidden indel creation rate
--length 150 30 length of reads in bp, must be >30
--mutation 0 0-1 hidden mutation rate
--stdev 50 standard deviation of --distance
--lr-type -l haplotagging type of linked-read experiment
--molecule-coverage -c 0.2 mean percent coverage per molecule if <1, else mean number of reads per molecule
--molecule-length -m 80000 mean length of molecules in bp
--molecule-number -n 3 mean number of unrelated molecules per barcode, where a negative number (e.g. -2) will use a fixed number of unrelated molecules and a positive one will draw from a Poisson distribution

# Barcodes

# Randomly generate

Mimick lets you put in length and count parameters, which it will use to randomly generate barcodes. The format is length,count (no spaces), where length is the base-pair length the barcodes should be and count is how many barcodes it should generate of length length. For example, if you specify 16,4000000, Mimick will generate 4 million unique 16bp barcodes, effectively mimicking 10X barcodes. Mimick will write a file containing the barcodes it generated. In practice, this would look something like:

randomly generated barcodes
harpy simulate linkedreads --lr-type 10x 16,4000000 hap1.fasta hap2.fasta
#                                     16bp^ ^4 million barcodes

# Specific barcodes

Alternatively, if you have a set of barcodes you absolutely want to use, just put the filename as the first positional argument. In practice, this would look something like:

barcodes as a file
harpy simulate linkedreads --lr-type 10x barcodes.txt hap1.fasta hap2.fasta
#                                        ^file of barcodes

# FASTA file(s)

You will need at least 1 fasta file as input, which goes at the very end of the command. It's assumed that each fasta file is a different haplotype of the same genome.

fasta inputs
harpy simulate linkedreads --lr-type 10x 16,4000000 hap1.fasta hap2.fasta
#                                         ^barcodes  ^fasta 1   ^fasta 2
linked-read

© Copyright 2025. All rights reserved.