Intrinsic DNA properties such as bending play a crucial role in diverse biological systems. Recent advantage in high-throughput method called loop-seq makes it possible to determine bendability of hundred thousand 50-bp DNA duplexes in one experiment. However, it’s still infeasible to assess whole sequence bendability in large genomes such as human, which needs thousands loop-seq experiments. Here we introduce ‘bendnet’ – a neural network to accurately predict the intrinsic DNA bending at base-resolution by only given DNA sequences. Bendnet can increase the resolution of experimental results, and can predict DNA bendability for any new given sequences in high accuracy. We applied bendnet to the human genome and observed a high-stiffness region located at both transcriptional start sites and transcriptional end sites. Such stiffness patterns are different for coding and non-coding genes, which matches distinct nucleosome occupancy patterns. As we expected, most of the transcription factors (TFs) bind in DNA of low bendability. In contrast, we observe an unusually high bendability within binding elements of specific TFs such as EBF1 and regulators of genome folding such as CTCF. These factors either co-bind or compete with nucleosome to carry out their functions. More interestingly, CTCF binding regions exhibit the highest bendability than other DNA elements that may help to trap and hold the CTCF in the exact location implying how CTCF work as stable anchor in loop extrusion process. Our work provides a tool to assess DNA bendability for large scale DNA sequences and expands our understanding on DNA mechanics in chromatin regulation and genome folding.

DNA mechanics,transcriptional regulation