At a time when profound gaps remain in our understanding of how one person's genes differ from another person's, developing accurate methods to interpret the human genome is essential to uncover population history, population relationships and the genetic basis of some of the most common diseases affecting society. Recently published in PLoS Genetics, new research conducted by Kennedy Krieger Institute, in collaboration with software company Partek Inc., has uncovered a novel mathematical method that combines two commonly used genotype association approaches to reveal the often opaque genetic distinction between individuals.
"Our results come at a time when the research community is beginning to make huge strides in our understanding of the molecular pathways within our own genome," said Dr. Jonathan Pevsner, senior study author and Director of Bioinformatics at the Kennedy Krieger Institute. "What is so exciting is that our approach gives the research community a more accurate and improved method to evaluate genome-wide studies and improve the field of genomics research as we know it."
Currently, genome-wide association studies (GWAS) are conducted to examine genes of different individuals to determine genetic variation from individual to individual. In genetic disease studies, researchers are most interested in unrelated individuals with the same disease that share common alleles (versions of a gene). If two unrelated individuals with the same disease randomly have common alleles at a given chromosomal region, researchers can potentially link that region to the disease.
"With so many unanswered questions about the human genome, we need to continually look for new methods to more precisely define relationships," said Thomas Downey, study author and President of Partek Inc. "Identifying genetic biomarkers within the population will allow the research community to better understand the relationship between an inherited disease and its genetic cause."
A variety of tools are currently available to make sense of genetic data. However, Dr. Pevsner and his colleagues developed a new mathematical method, in collaboration with Partek Inc., to estimate the relatedness between populations by combining information about how genetic material is shared between any two people, and how it is transmitted from parents to their children. In contrast to the existing methods, using the new calculation, Dr. Pevsner's team was able to decipher the following without reliance on any prior data:
•Previously unknown familial relationships and population relatedness in broad genomic data
•The nature of relatedness between particular individuals
•Fewer false positives (defined as individuals who are unrelated, but who are called as related)
Using this information, researchers can narrow down the possible candidates for a given inherited disease by identifying multiple common single nucleotide polymorphisms (SNPs), or genetic variations in a single DNA sequence, within a haplotype (combination of alleles). Haplotype SNPs often serve as significant clinical biomarkers in drug discovery and development.
"Our results not only reveal new insights into human evolution, but also provide the research community with better tools to identify biomarkers for disease," said Dr. Pevsner. "With more precise biomarkers, we will be able to develop more targeted treatments against genetic diseases that afflict the population."
Notably, when testing their method on the Human Variation Panel, Dr. Pevsner and his colleagues found previously unidentified identical, parent-child, and full-sibling relationships. Studies that calculate the associations between alleles, like the ones above, rely on unbiased representative sampling of population subsets to ensure validity. Consequently, samples are biased when unreported familial relationships are present in the dataset, causing a more frequent occurrence of common SNPs and likely misinterpretation of data and potentially unreliable findings.
"Surprisingly, when applying our new method to the Human Variation Panel, we found previously undeclared relationships," said Dr. Pevsner. "Our findings have an immediate impact on researchers studying this collection and perhaps an even more widespread impact on any findings that were based on the Human Variation Panel."
Dr. Pevsner and his colleagues will continue to apply this method to other population subsets in hopes of further decoding the links between our genes and disease.
Gene editing with CRISPR may introduce new defects