Reconstructing human somatic cell fates with fluctuating CpG methylation

Darryl Shibata, MD, USC 
Professor of Pathology
 

Wednesday, May 6, 2026
1:30 - 2:30 PM
ChEM-H & Neurosciences Bldg., Rotunda E241 
 

Abstract:  

Normal and neoplastic human cell ancestries can be reconstructed from their somatic alterations (molecular clock hypothesis). A current limitation is that molecular clocks lack the temporal resolution to reconstruct events occurring over only a few months to years because their error rates (~10-9 for mutations) are low. One solution is fluctuating CpG (fCpG) methylation, where balanced oscillating methylation and demethylation error rates are as high as 10-2 to 10-3 per division---a clock with several hundred fCpG sites will reliably distinguish daughters from most mitosis. The fCpG clock can infer mitotic ages---a new clonal population will have a synchronized “W-shaped” distribution with 0, 50, and 100% peaks reflecting the pattern of its progenitor cell. fCpG methylation becomes randomized between cells, and drifts to unimodal distributions around 50% with more divisions. fCpG clock principles will be outline with normal (intestinal crypts, brain) and neoplastic tissues (leukemia, breast cancer). The ages and ancestries of human cells provide high-resolution evolutionary frameworks to help interpret the myriad of available high-dimensional physical biological measurements.