KAGE is a tool for efficiently genotyping short SNPs and indels from short genomic reads.
KAGE requires Python 3, and can be installed using Pip:
pip install kage-genotyper
Test that the installation worked:
kage test The above will perform genotyping on some dummy data and should finish without any errors.
KAGE is easy and fast to use once you have indexes built for the variants you want to genotype. However, building these indexes can take some time. Therefore, we have prebuilt indexes for 1000 Genomes Projects variants (allele frequency > 0.1%), which can be downloaded from here.
If you want to make your own indexes for your own reference genome and variants, you should use the KAGE Snakemake pipeline which can be found here. Feel free to contact us if you want help making these indexes for your desired variants.
Once you have an index of the variants you want to genotype, running KAGE is straight-forward:
kmer_mapper map -b index -f reads.fa -o kmer_counts -l 150The Kmer Mapper requires for now a two-line fasta file. If you have a fastq file, you should convert that to fasta before mapping (e.g. by using Seqtk).
Note: Make sure l is the max read length of your input reads, and not any lower than that. The index specified by -b is an index bundle (see explanation above).
Count kmers:
kage genotype -i index -c kmer_counts --average-coverage 15 -o genotypes.vcfMake sure to set --average-coverage to the expected average coverage of your input reads. The resulting predicted genotypes will be written to the file specified by -o.
Note:
KAGE puts data and arrays in shared memory to speed up computation. It automatically frees this memory when finished, but KAGE gets accidentally killed or stops before finishing, you might end up with allocated memory not being freed. You can free this memory by calling kage free_memory.