A tandem repeat (TR) catalog generated from high-quality long-read human genome assemblies

This repository keeps the analysis scripts that were used to generated the TR catalog from public diploid long-read human genome assemblies from the following data soucres:

1. Human Pangenome Reference Consortium (HPRC)
2. Human Genome Structural Variation Consortium (HGSVC2)
3. 1000G ONT Sequencing Consortium

Workflow

Mapping of TRs from assemblies to the reference genome

Catalog

v1 v2

• haplotype names separated by semi-colons are shown in first header line preceded by '#'
• column descriptions:

Column	Description
chrom	chromosome
start	start coordinate
end	end coordinate
motif	consensus repeat motif
copy_numbers	copy numbers in haplotypes separated by semi-colons ('-' for missing genotypes)
sizes	sizes (bp) in haplotypes separated by semi-colons ('-' for missing genotypes)
motifs	motifs in haplotypes separated by semi-colons ('-' for missing genotypes)
max_change	maximum change (of all haplotypes) in size (bp) substracted from reference genome size
num_samples	number of samples with genotype
num_calls	number of haplotypes with genotype
motif_frequency	number of haplotypes associated with each motif observed e.g. CAG(10);CAA(2)
feature	gene element overlapped. Format: gene|transcript|, where = exon#|intron#|utr5|utr3|cds|promoter|exon_bound (exon boundary)