RFMIX is a program to identify the ancestry of genomic segments using random forest discriminative machine learning methods combined with a conditional random field model of the linear chromosome.
Allocate an interactive session and run the program.
Sample session (user input in bold):
[user@biowulf]$ sinteractive salloc.exe: Pending job allocation 12345678 salloc.exe: job 12345678 queued and waiting for resources salloc.exe: job 12345678 has been allocated resources salloc.exe: Granted job allocation 12345678 salloc.exe: Waiting for resource configuration salloc.exe: Nodes cn1234 are ready for job [user@cn1234 ~]$ module load rfmix [user@cn1234 ~]$ rfmix --help RFMIX v2.03-r0 - Local Ancestry and Admixture Inference (c) 2016, 2017 Mark Koni Hamilton Wright Bustamante Lab - Stanford University School of Medicine Based on concepts developed in RFMIX v1 by Brian Keith Maples, et al. This version is licensed for non-commercial academic research use only For commercial licensing, please contact cdbadmin@stanford.edu --- For use in scientific publications please cite original publication --- Brian Maples, Simon Gravel, Eimear E. Kenny, and Carlos D. Bustamante (2013). RFMix: A Discriminative Modeling Approach for Rapid and Robust Local-Ancestry Inference. Am. J. Hum. Genet. 93, 278-288 Summary of command line options - see manual for details -f, --query-file= (required) VCF file with samples to analyze (required) -r , --reference-file= (required) VCF file with reference individuals (required) -m , --sample-map= (required) Reference panel sample population classification map (required) -g , --genetic-map= (required) Genetic map file (required) -o , --output-basename= (required) Basename (prefix) for output files (required) --chromosome= (required) Execute only on specified chromosome (required) -c , --crf-spacing= Conditional Random Field spacing (# of SNPs) -s , --rf-window-size= Random forest window size (class estimation window size) -w , --crf-weight= Weight of observation term relative to transition term in conditional random field -G , --generations= Average number of generations since expected admixture -e , --em-iterations= Maximum number of EM iterations --reanalyze-reference In EM, analyze local ancestry of the reference panel and reclassify it -n , --node-size= Terminal node size for random forest trees -t , --trees= Number of tree in random forest to estimate population class probability --max-missing= Maximum proportion of missing data allowed to include a SNP -b , --bootstrap-mode= Specify random forest bootstrap mode as integer code (see manual) --rf-minimum-snps= With genetic sized rf windows, include at least this many SNPs regardless of span --analyze-range= Physical position range, specified as - , in Mbp (decimal allowed) --debug= Turn on any debugging output --n-threads= Force number of simultaneous thread for parallel execution --random-seed= Seed value for random number generation (integer) (maybe specified in hexadecimal by preceeding with 0x), or the string "clock" to seed with the current system time. [user@cn1234 ~]$ vartrix --bam test_dna.bam --cell-barcodes dna_barcodes.tsv --fasta test_dna.fa --out-matrix test_dna.out --vcf test_dna.vcf [user@cn1234 ~]$ exit salloc.exe: Relinquishing job allocation 12345678 [user@biowulf ~]$
Create a batch input file (e.g. vartrix_job.sh). For example:
#!/bin/bash set -e module load vartrix vartrix --bam test_dna.bam \ --cell-barcodes dna_barcodes.tsv \ --fasta test_dna.fa \ --out-matrix test_dna.out \ --vcf test_dna.vcf
Submit this job using the Slurm sbatch command.
sbatch [--cpus-per-task=#] [--mem=#] vartrix_job.sh
Create a swarmfile (e.g. vartrix_jobs.swarm). For example:
vartrix --bam test_dna1.bam \ --cell-barcodes dna_barcodes1.tsv \ --fasta test_dna1.fa \ --out-matrix test_dna1.out \ --vcf test_dna1.vcf vartrix --bam test_dna2.bam \ --cell-barcodes dna_barcodes2.tsv \ --fasta test_dna2.fa \ --out-matrix test_dna2.out \ --vcf test_dna2.vcf vartrix --bam test_dna3.bam \ --cell-barcodes dna_barcodes3.tsv \ --fasta test_dna3.fa \ --out-matrix test_dna3.out \ --vcf test_dna3.vcf
Submit this job using the swarm command.
swarm -f vartrix_jobs.swarm [-g #] [-t #] --module vartrixwhere
| -g # | Number of Gigabytes of memory required for each process (1 line in the swarm command file) |
| -t # | Number of threads/CPUs required for each process (1 line in the swarm command file). |
| --module vartrix | Loads the vartrix module for each subjob in the swarm |