Cell paper: FIRE combines machine learning and DNA sequencing to pinpoint gene regulatory switches

Mitchell R. Vollger and colleagues developed FIRE, described in Cell, which combines machine learning with DNA sequencing to locate regulatory differences between maternally and paternally inherited chromosomes. The team found that in about 40% of cases, regulatory differences between maternal and paternal copies occur at sites with identical DNA sequences. FIRE also located a gene-switch variant linked to celiac disease risk that disrupts a binding site for gene-control proteins, and identified tens of thousands of candidate regulatory regions; the paper's DOI is 10.1016/j.cell.2026.09.034.