NanoLoop: A deep learning framework leveraging Nanopore sequencing for chromatin loop prediction

Chromatin loops play a crucial role in gene regulation and cellular function, providing key insights into understanding the three-dimensional structure of the genome and its impact on cellular homeostasis. Nanopore sequencing technology, with its advantages in simultaneously detecting sequences and methylation patterns, brings new opportunities for studying three-dimensional genome structures. We introduce NanoLoop, the first algorithmic framework attempting to predict genome-wide chromatin interactions using Nanopore data. In experiments across four human lymphoblastoid cell lines, NanoLoop demonstrated excellent predictive performance and cross-cell line generalization capabilities. We also discovered four distinct methylation patterns at loop anchors that influence histone modification levels and determine various loop types. NanoLoop further predicted previously uncharacterized long-range chromatin loops, highlighting DNA methylation’s role in three-dimensional genome regulation and providing new insights into the complex regulatory relationships between epigenetic modifications and three-dimensional genome organization.