The plasma of individuals contains cell-free DNA (cfDNA) fragments released by various tissues in the body. These cfDNA fragments possess molecular characteristics, such as DNA methylation status, that reflect their cellular origins. The fragmentation of plasma DNA is influenced by factors such as nucleosomal organization, chromatin structure, gene expression, and nuclease content in the tissue of origin. As a result, characteristic signatures in the form of fragment size, nucleotide motifs at the fragment ends, single-stranded jagged ends, and the genomic locations of the fragmentation endpoints are observed. 

Professor Lo's research team discovered that these cfDNA fragment patterns contain rich genetic and epigenetic information and are not randomly fragmented. Expanding on this finding, Professor Lo's team published a paper titled "Epigenetic analysis of cell-free DNA by fragmentomic profiling" in PNAS in 2022. This research overcame the limitations of bisulfite sequencing and demonstrated the potential to infer cfDNA methylation patterns from cfDNA fragment patterns. By analyzing the cleavage profiles of cfDNA around cytosine-phosphate-guanine (CpG) sites and utilizing deep learning algorithms, the team successfully determined the methylation status from specific regions to individual CpGs. This confirmed that genetic and epigenetic information in cfDNA can be obtained through a single non-invasive testing method. The team elucidated the feasibility of using cfDNA cleavage patterns to deduce CpG methylation at single CpG resolution using a deep learning algorithm and achieved an AUC of 0.93 This fragmentomic-based methylation analysis (FRAGMA) opens up new possibilities for non-invasive prenatal screening, cancer detection, and organ transplant evaluation. Professor Lo's research in this area has deepened our understanding of the biology and generation of cell-free DNA, shedding light on the roles of nucleases in plasma DNA biology. This comprehensive approach to analyzing cell-free DNA fragments has the potential to impact clinical practice by enabling targeted DNA methylation, fragmentomic, and topologic analyses on a genome-wide scale. This advancement holds promise for improving early cancer detection and guiding personalized treatment approaches.