Characterization of missing human genome sequences and copy-number polymorphic insertions

Abstract

The extent of human genomic structural variation suggests that there must be portions of the genome yet to be discovered, annotated and characterized at the sequence level. We present a resource and analysis of 2,363 novel insertion sequences corresponding to 720 genomic loci. We show that a substantial fraction of these sequences are either missing, fragmented or mis-assigned when compared to recent de novo sequence assemblies from short-read next-generation sequence data. We determine that 18–37% of these novel insertions are copy-number polymorphic, including loci that show extensive population stratification among Europeans, Asians and Africans. Complete sequencing of 156 of these insertions identifies novel exons and conserved non-coding sequences not yet represented in the reference genome. We develop a method to accurately genotype these novel insertions by mapping next-generation sequencing datasets to the breakpoint thereby providing a means to characterize copy-number status for regions previously inaccessible to SNP microarrays.

The immunoglobulin heavy chain locus: genetic variation, missing data, and implications for human disease

“…91 A portion of these described haplotypes were also shown to represent 'common' alleles, highlighting the fact that the current human reference assembly represents the minor allele for many loci across the genome; a point that is particularly important if the assembly sequence represents an allelic deletion. 91 With respect to IGHV, it is interesting to point out that many of the identified alternate haplotypes known in the cluster also represent variants that are at higher frequencies than those found in the reference genome. 30,35 The recent development of whole-genome NGS and assembly has expanded our ability to identify and genotype SNPs and structural variants (insertions, deletions and inversions).…”

“…Extensive efforts have been undertaken in recent years to better characterize genomic structural variants in such loci. [89][90][91] Many of these have been discovered and sequenced by fosmid-end mapping and the generation of complete sequence from 'discordant' clones, relative to the reference genome assembly.…”

“…Indeed, targeted genotyping using NGS approaches based on data from completely sequenced alternate haplotypes has been shown to be an effective method for assaying polymorphisms in other regions of the genome. 91 The use of targeted fosmid-based sequencing in the IGHV gene region for describing alternate CNV-containing haplotypes would likely prove effective, as such methods allow for directed sequence assembly of single 40 Kb haploid segments, reducing discrepancies that arise when attempting to discern allelic SNPs from sequence differences between duplicated sequence blocks. An additional advantage is that existing fosmid-based resources 90,91 have been generated from DNA samples collected across a diverse range of Figure 5.…”

Structural variation in two human genomes mapped at single-nucleotide resolution by whole genome de novo assembly

“…Each pool consisted of only 30 fosmids to avoid ambiguities owing to segmented duplications in the assembly process. And each pool was sequenced independently to reduce the complexity of assembling the reads and problems caused by repetitive sequences 41 . Using the contigs assembled from fosmids, we validated 11 (73%) of the remaining 15 structural variations, which indicates that failure to amplify a variant by PCR is not necessarily indicative of a spurious structural variation call (Supplementary Table 2).…”

A Biologically Informed Method for Detecting Associations with Rare Variants

“…Since the reference genome is predominantly of European ancestry [98,99,100], populations with non-European ancestry generally have more variation with respect to the reference genome than populations of European ancestry (see Table 14). Therefore, to interpret the results of this study, one might conclude that non-European populations have higher rates of sequencing error than European descent populations.…”