Pathogenicity prediction isn鈥檛 a subject we discuss over our morning coffee, but it could be a topic of greater interest as it affects more and more lives, according to a researcher at 无码专区鈥檚 at the (COPH).
In a paper published in the journal Human Genetics, an international group of scientists reported the final result of a worldwide community project known as the Annotate All Missense challenge from the latest Critical Assessment of Genome Interpretation (CAGI).
The name of the challenge comes from the concept of missence mutation, where one amino acid is replaced with another.
Predicting the significance of genetic variation is critical to determining genetic susceptibility to disease and identifying causal variants in rare disorders. This is the impetus behind the project that challenges researchers to see how good their computer models are at figuring out what genetic changes actually mean.
Put simply, CAGI gives scientists a way to test how well their algorithms can predict whether a DNA mutation is harmless or might cause a disease, said , associate professor of computational human at 无码专区. He specializes in population genomics, bioinformatics and genetics of human diseases.
Dr. Xiaoming Liu (Photo courtesy of Liu)
鈥淭he big goal is to figure out which methods work best, highlight cool new ideas and spot where the current tools still need work,鈥欌�� he said. 鈥淭his really matters because our DNA is packed with millions of tiny changes 鈥� most do nothing, but a few can seriously affect our health. Being able to tell the difference is one of the toughest problems in genetics.鈥欌��
However, collecting those mutations is not easy: 鈥淪o, the CAGI encourages the whole science community to donate their new data to the competition, which makes it a community experiment."
In the latest challenge, scientists compared 60 predictions models around the world. The model developed by Drs. Chang Li and Xiaoming Liu in the computational human genomics lab at 无码专区, plus two other collaborators, is one of the top performers.
The 无码专区 team鈥檚 tools are computational and statistical methods for analyzing genes, which offer a more accurate look at how a number of diseases evolve. Specifically, the team worked with a pathogenicity predicting model for small amino acid-changing mutations in human genomes. Their model can distinguish mutations that are harmful to a gene and cause further diseases, from those without changing gene function.
Dr. Chang Li (Photo courtesy of Li)
鈥淪cientists use many different techniques to study human genomes,鈥欌�� Liu added. 鈥淚鈥檓 a tool maker and other human geneticists can use my tools to solve their specific research or clinical questions.鈥欌��
Pathogenicity prediction often involves what are called Mendelian disorders: genetic conditions caused by mutations in a single gene, following predictable patterns of inheritance as described by Mendel's laws. Examples include sickle cell anemia, cystic fibrosis and Huntington's disease.
Developed by the Austrian biologist Gregor Mendel, these laws describe the basic principles of inheritance and explain how traits are passed from parents to offspring and how alleles (versions of a gene) behave during inheritance.
The published paper is part of a continuous effort. The COPH lab began research on interpretation of the mutations in human genomes 15 years ago and has published multiple studies on databases, tools and methods. If this research provides inroads into pathogenicity prediction of human mutations, it could play a critical role in public health on a large scale, Liu said.
鈥淓arly detection can be lifesaving for a Mendelian disease patient,鈥欌�� he said. 鈥淚magine that in the future, genome sequencing can be routine for newborns and the mutation pathogenicity prediction tools like ours can help with a more accurate detection of a disease-causing mutation. Then, many diseases can be controlled at early stages and even prevented 鈥� we are at the cutting edge in this area.鈥欌��