Personal genome sequencing: current approaches and challenges

Abstract

The revolution in DNA sequencing technologies has now made it feasible to determine the genome sequences of many individuals; i.e., "personal genomes." Genome sequences of cells and tissues from both normal and disease states have been determined. Using current approaches, whole human genome sequences are not typically assembled and determined de novo, but, instead, variations relative to a reference sequence are identified. We discuss the current state of personal genome sequencing, the main steps involved in determining a genome sequence (i.e., identifying single-nucleotide polymorphisms [SNPs] and structural variations [SVs], assembling new sequences, and phasing haplotypes), and the challenges and performance metrics for evaluating the accuracy of the reconstruction. Finally, we consider the possible individual and societal benefits of personal genome sequences.

Figure 1.

Cost of DNA sequencing and cumulative number of genomes sequenced as a function of time. The blue points and the fitted line show the per-base sequencing cost, and the red points show the total number of sequenced genomes.

Figure 2.

Types of variations present in a human genome sequence. The reference genome is shown at the top of each subfigure, with the individual's diploid genome shown below it. (A) A heterozygous insertion SNP and a homozygous small deletion. (B) Two phased SNPs. (C) A homozygous deletion in the target genome. (D) A heterozygous novel insertion. (E) A heterozygous inversion event.

Figure 3.

Length distribution of Indels. The figure (in log–log scale) shows the size distribution of all of the homozygous and heterozygous Indels in the HuRef genome (Levy et al. 2007) as compared with the NCBI reference.

Figure 4.

Methods for detecting variation in a human genome sequence using DNA sequencing technologies. (A) Paired-end reads to detect insertions and deletions. (B) Split read methods for breakpoint idnetification. (C) Read depth analysis to detect CNVs. (D) Local reassembly to reconstruction novel insertions.

Figure 5.

A general flow chart for determining a personal genome sequence. The generated reads are first mapped to the reference genome to call high-quality SNPs and small Indels (Step 1) and SVs based on aberrant alignment information (Step 2). The novel insertions can be reconstructed using local de novo assembly algorithms (Step 3), and a final phasing step (Step 4) will be able to deduce the complete diploid genome of the individual.