Genomes

GenomeKit now supports building and using assemblies and annotations. For assemblies, the schema will follow the UCSC format, and for annotations, they must be specified in GENCODE/Ensembl/NCBI GFF3 formats.

Examples

starter/build.sh is a bash script that builds a few assemblies and annotations to get you started.

Assemblies

1. Copy the 2bit, chrom.sizes, and chromAlias.txt files from https://hgdownload.soe.ucsc.edu/goldenPath/hg38/bigZips/. into your

   python -c 'import appdirs; print(appdirs.user_data_dir("genome_kit"))'
   

   directory.

   If you need to generate from a fasta:

   1. conda create -n ucsc-tools ucsc-fatotwobit ucsc-twobitinfo
      

   2. conda activate ucsc-tools
      

   3. follow the instructions at https://genome.ucsc.edu/goldenPath/help/twoBit.html

   4. optionally create an chromAlias.txt with any contig aliases required.

Annotations

1. python -c 'import genome_kit as gk; print(gk.GenomeAnnotation.build_gencode("MY_ANNO.gff3", "MY_ANNO", gk.Genome("MY_ASSEMBLY")))'
   

2. Copy the resulting files into the

   python -c 'import appdirs; print(appdirs.user_data_dir("genome_kit"))'
   

   directory.

   The .dganno file contains the compiled GFF3 and the .cfg file contains metadata, such as refg=hg38.