Insilico Study Identifies Targets in Rare Sinonasal Cancer

The researchers compared normal sinonasal epithelium, inverted papilloma, and carcinoma using whole-exome, RNA, and mitochondrial DNA sequencing. They reported a stepwise change in gene activity across those disease stages, including increased cell-cycle, extracellular-matrix, and metabolic signaling alongside reduced immune and apoptotic signaling. Shared genomic changes appeared in only some paired papilloma and carcinoma samples, while the mitochondrial analysis found no overlapping mutations.

The team then applied Insilico's PandaOmics platform to genes whose expression rose during disease progression. The first pass prioritized targets with existing FDA-approved inhibitors and identified CDK6, EGFR, HDAC, and SRC/YES1 as drug-repurposing candidates. A separate set—AURKA, PLK4, TTK, and CDK1/7—was nominated for preclinical investigation.

The work involved researchers affiliated with the University of Chicago, Johns Hopkins University, and Insilico Medicine. Nature's article record and the PubMed record both identify July 10, 2026, as the paper's publication date.

These are therapeutic hypotheses, not demonstrated treatments for patients. The study did not establish clinical efficacy, and its 11-patient cohort limits how broadly the molecular patterns can be generalized. The ranked targets require functional validation in disease models and, if supported there, further translational and clinical testing.

For computational drug-discovery teams, the study illustrates a bounded role for target-ranking software: molecular measurements established the disease-stage signals first, and the AI platform then helped narrow a large candidate space. That sequencing matters because target priority is not the same as biological validation. In rare cancers where samples are scarce, such workflows can organize follow-up experiments, but they cannot remove the need for independent replication or clinical evidence.