If you’re investigating structural variations underlying cancer-causing genes, you’ll want to see this new algorithm called “HYENA” (or, “Hijacking of Enhancer Activity”).

True to its unique name, HYENA is proving to be an effective hunter—helping to track structural variations that shuffle DNA from one location to another to boost overexpression of cancer-causing genes.

According to the authors, HYENA offers a gene-centric approach to help you search for elevated gene expression. What sets HYENA apart from other models is that it identifies both coding and noncoding genes using whole-genome and transcriptome sequencing data.

Using HYENA, researchers found 108 candidate oncogenes, including many non-coding genes, in more than 1,000 adult tumors from 25 tumor types. Most of these candidate genes link to pancreatic cancer.

Corresponding author, Dr. Lixing Yang, of the University of Chicago, noted, “Although we don’t yet have a ‘gold standard’ available to comprehensively evaluate HYENA’s performance, we were able to compare it to two existing algorithms. We found that our model was superior in both sensitivity and specificity in detecting enhancer hijacking genes. Most importantly, unlike the other algorithms, HYENA helped us identify even non-coding targets.”

He added, “It’s possible that some of the candidate genes we found are not oncogenes. Our next steps will be to conduct functional studies to determine the disease relevance of these candidate genes.”

NCI’s Division of Cancer Biology funded this work.