GTF preparation

  1. Obtain a GTF file for the reference genome with which your reads are aligned.
  2. Convert GTF to reduced GTF using Drop-seq ReduceGTF. You’ll need a sequence dictionary for the reference genome. You can create one using Picard CreateSequenceDictionary or another tool.
    You can edit your sequence dictionary so that it only contains the X contig to make the reduced GTF smaller (optional).
  3. Prepare the GTF for xipher. 
    xipher::prepareGtfClp(outAnnotationsPath, inReducedGtfPath)

    GnomAD preparation (optional)

    TODO: sburger explain why to do this

  4. Download the genome chrX gnomAD VCF file. See gnomAD downloads for more information.
  5. Run GATK VariantsToTable to convert the VCF to a table. 
    java -jar GenomeAnalysisTK.jar \
VariantsToTable \
-R your.fasta \
-V gnomad.genomes.v?.sites.chrX.vcf.bgz \
-F CHROM -F POS -F REF -F ALT -F TYPE \
-F AC -F AN -F AF \
-O gnomad_chrX_variants_table.txt
    • your.fasta is the reference genome fasta file you used to align your reads.
    • gnomad.genomes.v?.sites.chrX.vcf.bgz is the gnomAD VCF file you downloaded.
    • gnomad_chrX_variants_table.txt is the output file.
  6. Prepare the gnomAD table for xipher. 
    xipher::prepareGnomAdClp(outGnomADPath, inGnomADVariantsTablePath)
    • outGnomADPath is the path to the output file.
    • inGnomADVariantsTablePath is the path to the table created by GATK VariantsToTable.