Xudong Liu1,7, Ying Ni2,3,4,7, Lianwei Ye1,7, Zhihao Guo1, Lu Tan1, Jun Li1, Mengsu Yang2,3,4,5, Sheng Chen6 and Runsheng Li1,3,4 1Department of Infectious Diseases and Public Health, Jockey Club College of Veterinary Medicine and Life Sciences, City University of Hong Kong, Hong Kong 999077, China; 2Department of Biomedical Sciences, Jockey Club College of Veterinary Medicine and Life Sciences, City University of Hong Kong, Hong Kong 999077, China; 3Department of Precision Diagnostic and Therapeutic Technology, City University of Hong Kong Shenzhen Futian Research Institute, Shenzhen 518000, China; 4Tung Biomedical Sciences Centre, City University of Hong Kong, Hong Kong 999077, China; 5Key Laboratory of Biochip Technology, Biotech and Health Centre, Shenzhen Research Institute of City University of Hong Kong, Shenzhen 518000, China; 6State Key Lab of Chemical Biology and Drug Discovery, Department of Food Science and Nutrition, The Hong Kong Polytechnic University, Hong Kong 999077, China

↵7 These authors contributed equally to this work.

Corresponding authors: sheng.chenpolyu.edu.hk; runshelicityu.edu.hk Abstract

DNA modifications in bacteria present diverse types and distributions, playing crucial functional roles. Current methods for detecting bacterial DNA modifications via nanopore sequencing typically involve comparing raw current signals to a methylation-free control. In this study, we found that bacterial DNA modification induces errors in nanopore reads. And these errors are found only in one strand but not the other, showing a strand-specific bias. Leveraging this discovery, we developed Hammerhead, a pioneering pipeline designed for de novo methylation discovery that circumvents the necessity of raw signal inference and a methylation-free control. The majority (14 out of 16) of the identified motifs can be validated by raw signal comparison methods or by identifying corresponding methyltransferases in bacteria. Additionally, we included a novel polishing strategy employing duplex reads to correct modification-induced errors in bacterial genome assemblies, achieving a reduction of over 85% in such errors. In summary, Hammerhead enables users to effectively locate bacterial DNA methylation sites from nanopore FASTQ/FASTA reads, thus holds promise as a routine pipeline for a wide range of nanopore sequencing applications, such as genome assembly, metagenomic binning, decontaminating eukaryotic genome assemblies, and functional analysis for DNA modifications.

Received January 30, 2024. Accepted September 25, 2024.