Patterns of low-complexity regions in human genes

Abstract Genome evolution stands as a paramount determinant for species survival and overall biodiversity on Earth. Among the myriad processes orchestrating genome evolution, the dynamic attributes of length and compositional polymorphism within low-complexity regions (LCR) are the fastest. Clusters of LCR hotspots serve as pivotal conduits connecting different modes of genome evolution, specifically arising through gene duplication events and harboring pivotal sites susceptible to point mutations. Thus, they offer a holistic perspective on the panorama of genome evolution. Furthermore, LCR actively participates in a multifaceted spectrum of neurological, developmental, and cognitive disorders. Despite the substantial body of knowledge concerning the roles of individual LCR-containing genes in the causation of diseases, a comprehensive framework remains conspicuously absent, failing to provide a unified portrayal of LCR-containing genes and their interactions. Furthermore, our understanding of the intricate interplay between paralogy and LCR remains notably deficient. Within this study, we have identified nine clusters of LCR hotspots within the human genome. These clusters are predominantly comprised of closely positioned paralogs, characterized by a significantly higher prevalence of shared LCR and a lower degree of differentiation (F ST ) across diverse human populations. Moreover, we have unveiled intricate networks of LCR-containing genes engaged in mutual interactions, sharing associations with a spectrum of diseases and disorders, with a particular emphasis on hereditary cancer-predisposing syndromes. Our discoveries shed light on the compelling potential of LCR-containing interacting genes to collectively engender identical diseases or disorders, thereby underscoring their pivotal role in the manifestation of pathological conditions. Significance Statement Among myriad genome evolution processes, low-complexity regions (LCR) are pivotal, being both the fastest and bridging other evolution modes like gene duplication and point mutations. Understanding LCR-containing paralogous genes is essential to comprehend genetic diseases. Here, we demonstrate that the human genome harbors clusters of LCR hotspots mainly composed of paralogous genes sharing LCR, indicating a role for segmental duplication. The degree of differentiation is significantly lower in clusters of LCR hotspots than in other regions. Moreover, we provide a detailed network of LCR-containing interacting genes associated with shared diseases. Instead of attributing a single disease to an LCR gene, a unified perspective on LCR-containing interacting genes causing the same disease enhances our understanding of LCR-induced disease mechanisms.