Classification and monomer-by-monomer annotation of suprachromosomal family 1 alpha satellite higher-order repeats in hg38 human genome assembly

Abstract

In the latest hg38 human genome assembly, centromeric gaps has been filled in by alpha satellite (AS) reference models (RMs) which are statistical representations of homogeneous higher-order repeat (HOR) arrays that make up the bulk of the centromeric regions. We studied these models to compose an atlas of human HORs where each monomer of a HOR could be characterized and represented by a number of its polymorphic sequence variants. We further used these data and HMMER sequence analysis platform to annotate AS HORs in the assembly. This led to discovery and annotation of a new type of low copy number highly divergent HORs which were not represented by RMs. The annotation can be viewed as UCSC Genome Browser custom track (the HOR-track) and used together with our previous annotation of AS SFs in the same assembly where each AS monomer can be viewed in its genomic context together with its classification into one of the 5 major SFs (the SF-track). To catalog the diversity of AS HORs in the human genome we introduced a new naming system. Each HOR received a name which showed its SF, chromosomal location and index number. Here we present the first installment of the HOR-track covering only the 17 HORs that belong to SF1 which forms live functional centromeres in chromosomes 1, 3, 5, 6, 7, 10, 12, 16 and 19 and also a large number of minor dead HOR domains, both homogeneous (pseudo) and divergent (relic). The 4 newly discovered divergent SF1 HORs have provided the missing links in SF1 early evolution and substantiated its partition into 2 generations, archaic and modern, which we reported earlier. Additionally, we demonstrated that monomer-by-monomer HOR annotation was useful for mapping and quantification of various structural variants of AS HORs which would be important for studies of inter-individual polymorphism of AS including centromeric functional epialleles.