> "a statistical analysis of data from 121 individuals possessing likely disease-causing variants of the gene. The team discovered that 25 of these individuals had a diagnosed mental disorder, including psychotic, mood, anxiety, personality, or eating disorders. Notably, all but two of these subjects carried a “null” variant of the gene, suggesting it was non-functional."

Garbage. Obviously p-hacked or of dubious significance, sample size isn't nearly large enough, and "mental illness" is not a coherent nor psychometrically valid category.