New technology may improve predicting cancer risk

Researchers at MD Anderson Cancer Center, in collaboration with the University of Utah and other institutions, have developed a powerful tool that combines two standard statistical methods of gene sequencing to more quickly identify mutations that cause disease.

In the study, published in May’s Nature Biotechnology, pVAAST (pedigree Variant Annotation, Analysis and Search Tool) was used to analyze the DNA of patients, their relatives and healthy people to predict the probability of the specific genetic variants and genes that are increasing the risk of developing disease.

“The tool uses family sequence data to find mutations responsible for both common complex diseases — such as common cancers — and rare Mendelian diseases, such as congenital heart defects,” said Chad Huff, Ph.D., assistant professor in Epidemiology at MD Anderson and corresponding author on the study.

This technology combines linkage analysis with case control association to help researchers and clinicians identify disease-causing mutations in families at a more accelerated rate than previous methods. Linkage analysis is used to track the shared genetic variants in families and synch the patterns with the disease symptoms. Case control association studies compared unrelated individuals with a specific disease to those who are healthy to search for genetic variants that are more common in one group than the other.

“Both methods were initially designed for genotype data, before the advent of affordable, high-throughput DNA sequencing (which produce thousands or millions of sequences concurrently),” said Huff. “The novelty of what we have done with pVAAST is to combine both approaches in a unified, automated framework for faster and more accurate disease-gene identification.”

For this study, pVAAST analyzed data to identify the genetic causes of three diseases: enteropathy (a chronic inflammation of the intestine), cardiac septal defects and Miller syndrome (a developmental defect of the face and multiple limbs). The new technology was able to identify the exact mutations causing these diseases from DNA data from a single family. In the cardiac septal defects and Miller syndrome families, the casual mutations previously had been identified and the results served as a proof of concept. In the enteropathy family, the causal mutation in the family was unknown prior to the analysis.

An added component of pVAAST is functional variant prioritization, which provides a computational prediction of whether a particular variant is likely to be damaging. The mathematical model in pVAAST also searches for de novo mutations that may be causing disease but are not present in the parents of affected individuals.

“We’ve had very effective tools for identifying rare, highly penetrant mutations,” said Huff. “Over the past nine years, we’ve also been able to use high-density genotyping to effectively identify common variants that increase the risk of cancer. However, until the last few years, it has not been possible to search the genome for rare variants that provide an intermediate increased risk of developing specific cancers.”

Huff says this is now possible, given the rapid reduction in the cost of whole exome and whole genome sequencing.

“Once these mutations are validated and characterized, clinical labs will be able to design genetic tests that provide better predictions of your personal risk of developing cancer,” said Huff.