hapLOH

hapLOH detects allelic imbalance and copy-neutral loss of heterozygosity (LOH) in heterogeneous tumor samples by integrating germline haplotype estimates with SNP array allele-specific intensity data.

Key Features:

• Enhanced sensitivity: Detects allelic imbalances at very low aberrant cell proportions, reportedly down to 0.25% using a 5 million SNP array.
• Haplotype-informed allele-specific intensities: Leverages germline haplotype estimates to interpret allele-specific intensity ratios for more precise detection of subtle allelic imbalance in unpaired tumor samples.
• Detection of somatic events: Identifies somatic segmental copy number variations and copy-neutral LOH events.
• Hidden Markov Model (HMM): Uses an HMM framework to segment the genome and call copy-neutral LOH, enabling detection of LOH segments as short as 11 megabases in mixtures containing ~4% tumor cells.
• Unpaired tumor sample analysis: Operates on unpaired tumor samples by exploiting germline haplotype information rather than requiring matched normal samples.
• Allele-specific intensity ratio analysis: Analyzes allele-specific intensity ratios from SNP microarrays as the primary signal for allelic imbalance detection.
• Validation on real and simulated data: Validated using real and simulated datasets, including the CRL-2324 breast cancer cell line genotyped on the Illumina 370K array.

Scientific Applications:

• Cancer genomics: Enables detection of low-frequency genomic aberrations in tumor samples for studies of tumor heterogeneity.
• Minimal residual disease monitoring: Supports identification of very low-abundance tumor cell populations that may indicate minimal residual disease.
• Tumor evolution studies: Facilitates analysis of clonal architecture and evolutionary dynamics by detecting subclonal allelic imbalances.
• Therapeutic response and resistance assessment: Assists investigation of genomic changes associated with treatment response and resistance mechanisms.

Methodology:

Analyzes SNP microarray allele-specific intensity ratios in the context of germline haplotypes and applies a hidden Markov model to segment and call copy-neutral LOH and allelic imbalance, with validation on real and simulated datasets including CRL-2324 genotyped on the Illumina 370K array.