A novel qualitative approach to analyse DNA methylation data from Deep Bisulfite Amplicon Sequencing (Deep- Bis)

DNA methylation is involved in a broad range of biological processes. Traditional quantitative approach summarises DNA methylation data as average percentage of methylated CpGs in specific genomic regions, or as methylation percentage for single CpG site. Such kind of approach gives important information on the general relationships between methylation and expression but might be not able to describe the complex epigenetic structure and dynamics within a cell population. Indeed, this approach fails to dissect and recognise different DNA methylation patterns when a heterogeneous population is investigated. In the present work, in order to better decode epigenetic data, a new way to analyse DNA methylation, based on qualitative approach, was developed. The qualitative approach allows methylation profiles of cell populations to be studied at the single molecule level, thus providing an added value to the quantitative one. In order to obtain an effective representation of methylation profiles, it has been used the Deep Bisulfite Amplicon Sequencing (Deep- Bis), which allows to obtain a very high coverage of selected loci. AmpliMethProfiler, a python-based pipeline, was developed to process the high number of sequences and to extract CpG methylation profiles at single molecule level from Deep- Bis data. Several tools borrowed from ecology and population genetics were used to describe the methylation landscape of the samples in terms of epialleles composition. Qualitative approach was applied on two experimental models: mouse development and acute myeloid leukemia (AML) progression before and after the demethylating therapy, in order to describe the methylation and demethylation dynamics, to investigate the epialleles distribution, to follow their evolution and to gain insight on epigenetic heterogeneity degree at specific loci. The tracking of the methylation profiles is more faithful to the epigenetic state of different loci and allows a more detailed overview of the methylation landscape in a tissue, which is composed by a mosaics of epigenetically different cells.