GIST Research Team Develops Algorithm for Detecting Genetic Mutations in Cancer Cells and Genome Restoration

[Asia Economy Honam Reporting Headquarters Reporter Lee Gwan-woo] Gwangju Institute of Science and Technology (GIST) announced on the 3rd that Professor Hyun-Joo Lee's research team from the Department of Electrical, Electronics and Computer Engineering developed a new graph-based algorithm that analyzes whole genome sequencing data to discover genetic variations and reconstruct genome structures at the single nucleotide level.

The research team succeeded in accurately discovering genetic variations from whole genome sequencing data, which typically has many variant detection errors, and identifying rearranged genome structures in cancer patients that were not detected by existing methods.

Whole Genome Sequencing data refers to data that provides the entire DNA base sequence of an individual organism.

The human genome consists of 3 billion base pairs, and cancer cells have genetic variations different from normal cells. Precisely identifying the distinct genetic variations according to each individual's cancer is important for personalized treatment.

However, analyzing 3 billion base pairs to accurately identify the variations present in cancer cells is a very challenging task. In particular, the rearranged chromosome structures observed in cancer cells under microscopes have never been identified at the single nucleotide level.

Therefore, an algorithm capable of analyzing whole genome sequencing data to identify these structures is needed.

The research team developed ‘InfoGenomeR’ (Integrative Framework for Genome Reconstruction), a genetic variation discovery and genome reconstruction algorithm. It converts sequences with structural variations into graph form, then reconstructs the graph so that structural variations and copy number variations have mutually consistent values, thereby reducing detection errors.

Subsequently, using heterozygous single nucleotide polymorphism information, they constructed a haplotype graph and restored the genome arrangement by finding an Eulerian path with the minimum entropy value.

InfoGenomeR developed by the research team significantly reduced genetic variation detection errors and reconstructed the genome arrangement of cancer cell lines at the single nucleotide level.

The accuracy of genetic variation detection was confirmed to be greatly improved compared to Manta, an algorithm by the international genome analysis company Illumina.

The research team applied the developed algorithm to breast and brain cancer patient data and identified that double minute circular genome structures were amplified dozens of times in patients, revealing that specific chromosomes undergo rearrangement processes depending on the cancer type.

They also discovered that rearranged genomes newly appeared in recurrent or metastatic cancers that were not present in the original cancer sites.

Professor Hyun-Joo Lee said, “Reconstructing the genome arrangement of cancer cells at the single nucleotide level using only whole genome sequencing data is a challenging problem and was not possible with existing algorithms, but InfoGenomeR is the first algorithm to successfully accomplish this.” He added, “We hope that based on the genome arrangement of cancer cells derived from this algorithm, we can elucidate the regulation of cancer-related gene expression for personalized medicine.”

This content was produced with the assistance of AI translation services.