GONE2¶
GONE2 estimates the recent history of effective population size (Ne) from the linkage disequilibrium between pairs of SNPs. It can also account for population structure, low-coverage sequencing data and genotyping errors.
License¶
Free to use and open source under GNU GPLv3.
Available¶
- Roihu-CPU: 2.0.2, via the
bio-appsmodule.
Usage¶
GONE2 is part of the bio-apps collection on Roihu. Load the bio-apps module tree and then the GONE2 module:
GONE2 reads genotype data in VCF, PED or TPED format. The file must be uncompressed
and its name must end in .vcf, .ped or .tped. GONE2 also needs the positions of
the SNPs on a genetic map. These are read from the .tped file itself, or, for .vcf
and .ped files, from a PLINK .map file with the same name in the same directory.
Alternatively, use -r to give a constant recombination rate in cM/Mb, and GONE2
converts the physical positions into a genetic map.
For example, analyse unphased diploid genotypes in data.vcf, assuming a
recombination rate of 1.1 cM/Mb across the genome:
Commonly used options:
-g: type of genotype data:0unphased diploid (default),1haploid,2phased diploid,3low-coverage-x: treat the sample as coming from a metapopulation with several subpopulations-sand-i: use a random subset of this many SNPs or individuals-t: number of threads-o: base name of the output files-S: seed for the random number generator. GONE2 uses a genetic algorithm, so results differ slightly between runs unless you fix both-Sand-t.
The Ne estimates for the last 150 generations are written to mypop_GONE2_Ne, the
observed and predicted linkage disequilibrium to mypop_GONE2_d2, and a summary of
the run to mypop_GONE2_STATS. With -x, the files are named mypop_GONE2_mix_*.
If you leave out -o, the output names start with the full input file name, for
example data.vcf_GONE2_Ne.
The module also provides gone2_tui, which runs the same analysis and shows its
progress and the Ne curve live in the terminal. Use it in an
interactive session, and gone2 in batch
jobs.
GONE2 on Roihu accepts at most 2,000,000 SNPs and 2,000 individuals (1,000 with
phased data, -g 2). For larger datasets, reduce your data first, for example with
PLINK. GONE2 always reserves about 4 GiB of memory for the genotype data,
whatever the size of your input, so reserve at least 8 GiB for the job.
Example batch script¶
#!/bin/bash
#SBATCH --job-name=gone2
#SBATCH --account=<project>
#SBATCH --output=output_%j.txt
#SBATCH --error=errors_%j.txt
#SBATCH --partition=small
#SBATCH --time=04:00:00
#SBATCH --nodes=1
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=8
#SBATCH --mem=8G
module load bio-apps/v202603
module load gone2/2.0.2
gone2 -r 1.1 -t $SLURM_CPUS_PER_TASK -o mypop data.vcf
Replace <project> with your CSC project (for example project_2001234), and 1.1
with a recombination rate suited to your species, or leave out -r and provide a
data.map file.
See creating a batch job script for Roihu for more information about running batch jobs.
Support¶
More information¶
- GONE2 GitHub repository
- Santiago, E., Köpke, C. & Caballero, A. (2025). Accounting for population structure and data quality in demographic inference with linkage disequilibrium methods. Nature Communications 16, 6054