 The paper proposes a novel approach to detecting 5-methylcytacine, 5-MC, in CPG, dinucleotides, CPGS. The authors use PAC by a circular consensus sequencing, CCS, to sequence a human sample and train a deep learning model called SMECSMETH. This model uses CCS reads to accurately identify 5-MC sites at single molecule resolution. Additionally, the authors developed a next-flow pipeline called SMETHPhase, which can detect haplotype-aware methylation using CCS reads. The authors tested their method on a Chinese family trio, and found that it was able to accurately detect 5-MC sites. This article was authored by Peng Nai, Fan Nai, Ziyu Chong, and others.