Tumor Haplotype Assembly Algorithms for Cancer Genomics

Abstract

The growing availability of inexpensive high-throughput sequence data is enabling researchers to sequence tumor populations within a single individual at high coverage. But, cancer genome sequence evolution and mutational phenomena like driver mutations and gene fusions are difficult to investigate without first reconstructing tumor haplotype sequences. Haplotype assembly of single individual tumor populations is an exceedingly difficult task complicated by tumor haplotype heterogeneity, tumor or normal cell sequence contamination, polyploidy, and complex patterns of variation. While computational and experimental haplotype phasing of diploid genomes has seen much progress in recent years, haplotype assembly in cancer genomes remains uncharted territory.In this work, we describe HapCompass-Tumor a computational modeling and algorithmic framework for haplotype assembly of copy number variable cancer genomes containing haplotypes at different frequencies and complex variation. We extend our polyploid haplotype assembly model and present novel algorithms for (1) complex variations, including copy number changes, as varying numbers of disjoint paths in an associated graph, (2) variable haplotype frequencies and contamination, and (3) computation of tumor haplotypes using simple cycles of the compass graph which constrain the space of haplotype assembly solutions. The model and algorithm are implemented in the software package HapCompass-Tumor which is available for download from http://www.brown.edu/Research/Istrail_Lab/.

Exploiting next-generation sequencing to solve the haplotyping puzzle in polyploids: a simulation study

“…The alignments are pre-processed to generate BAM files and remove duplicates by samtools [49] and Picardtools [50], after which SNPs are called using FreeBayes [21] (Figure 1-B). The processed alignments, the reference and the VCF files are used in the haplotyping step by HapCompass [39,43], HapTree [44] and SDhaP [40] to estimate the haplotypes using the phasing information from reads with at least two heterozygous SNPs (Figure 1-C). In the last step, the obtained estimates are compared to the original haplotypes by command-line tool hapcompare that we developed using several measures of estimation quality (Figure 1-D).…”

“…Here we review three state-of-the-art haplotyping algorithms applicable to polyploids: HapCompass [39,43], HapTree [44] and SDhaP [40], and evaluate their accuracy through extensive simulations of random genomes and NGS reads. Using the highly heterozygous tetraploid potato (S. tuberosum) as a model, we generated random genomes using a realistic stochastic model with parameters SNP density and distribution of SNP dosages, i.e.…”

Exploiting Next Generation Sequencing to Solve the Haplotyping Puzzle in Polyploids: A Simulation Study

“…The alignments are pre-processed to generate BAM files and remove duplicates by samtools [49] and Picardtools [50], after which SNPs are called using FreeBayes [21] (Figure 1-B). The processed alignments, the reference and the VCF files are used in the haplotyping step by HapCompass [39,43], HapTree [44] and SDhaP [40] to estimate the haplotypes using the phasing information from reads with at least two heterozygous SNPs (Figure 1-C). In the last step, the obtained estimates are compared to the original haplotypes by command-line tool hapcompare that we developed using several measures of estimation quality (Figure 1-D).…”

“…For simplicity, we focus on the most prevalent type SNPs, bi-allelic SNPs, for which the alleles can be represented by '0' (the reference) and '1' (the alternative). a) HapCompass: Aguiar and Istrail (2013) extend their graphical haplotype estimation approach for diploids [39], by constructing the polyploid Compass graph, which has k nodes for each variant site, s i , of a k-ploid corresponding to the k alleles at that site [43]. To each SNP pair, s i , s j , that is covered by at least one of the m fragments, the phasing with the largest likelihood is assigned by a polyploid likelihood model, conditional on the covering fragments and assuming a fixed base calling error rate.…”

CRISPR/Cas9‑induced saturated mutagenesis identifies Rad51 haplotype as a marker of PARP inhibitor sensitivity in breast cancer

“…In recent years, the CRISPR/Cas9-based negative screening strategy with high-throughput random screening has been applied to identify potential drug resistance mutations. Due to the limits of library construction, the induced mutation is identified within a limited region of the gene sequence, while randomly induced mutations are rarely observed in clinical samples and have limited clinical validity ( 52 – 54 ). The CRISPR/Cas9-based saturation mutation strategy to identify the drug resistance pathway hub gene displays high efficiency and clinical utility.…”