Nucleosome Footprint

The core mechanism relies on the protection of DNA wrapped around histone octamers. DNA segments tightly bound to nucleosomes are sterically shielded from endonuclease cleavage, while linker DNA between nucleosomes is exposed and preferentially cut. This differential protection generates a characteristic fragmentation pattern where cfDNA fragment ends map predominantly to linker regions, and the ~147 bp DNA wound around the histone core is enriched in sequencing libraries. The precise positioning of these nucleosomes acts as a latent epigenetic fingerprint.

Nucleosome positioning varies significantly between cell types, reflecting cell-specific gene regulatory landscapes. By analyzing the nucleosome-depleted regions (NDRs) at transcriptional start sites, algorithms can infer the contributing tissue sources in a mixed cfDNA sample. Key aspects include:

• Promoter Footprints: Active genes in the cell of origin exhibit open chromatin and NDRs at their promoters, producing distinct fragmentation profiles.
• Reference Maps: Deconvolution requires high-quality reference nucleosome maps from major contributing tissues (e.g., hematopoietic, hepatocyte, myocyte).
• Inference Algorithms: Non-negative matrix factorization or probabilistic mixture models estimate the fraction of cfDNA derived from each tissue.

Beyond nucleosome positioning, the footprinting signal can resolve sub-nucleosomal protection by transcription factors (TFs). When a TF is bound to its cognate motif, it transiently protects that short DNA segment from cleavage, leaving a small 'footprint' of increased read coverage flanked by accessible cut sites. This allows for the non-invasive monitoring of TF activity in the tissue of origin, providing a dynamic readout of signaling pathway activation that is complementary to static mutational analysis.

The diversity of fragment lengths around a genomic locus provides a quantitative metric of nucleosome phasing. Regions of ordered nucleosome arrays produce highly periodic fragment length distributions, while disordered or nucleosome-depleted regions exhibit high entropy. Metrics used include:

• Windowed Protection Score (WPS): A measure of the oscillation in coverage around a nucleosome dyad.
• Fragment Length Diversity: Shannon entropy calculated on the size distribution of fragments mapping to a specific window.
• Short-to-Long Ratio: The proportion of sub-nucleosomal (<150 bp) to mono-nucleosomal (150-250 bp) fragments, which increases at active regulatory elements.

Nucleosome footprinting provides orthogonal biomarker information to mutation-based liquid biopsy, particularly valuable in low tumor fraction settings or for cancers with few recurrent point mutations. Clinical utilities include:

• Cancer Detection: Global nucleosome disorganization in cancer genomes produces altered fragmentation profiles detectable even without identifying a specific driver mutation.
• Tissue of Origin Identification: Footprinting can localize a tumor to a specific organ by matching cfDNA patterns to tissue-specific chromatin maps, guiding the diagnostic workup for cancers of unknown primary.
• Treatment Response Monitoring: Dynamic changes in the footprinting signal at pharmacodynamic gene targets can indicate on-target drug activity in the tumor tissue.

Accurate footprinting requires rigorous correction for technical biases that can mimic or obscure biological fragmentation patterns. Critical normalization steps include:

• GC Bias Correction: Fragment end-motif preferences and sequencing efficiency vary with GC content, requiring loess regression or similar smoothing.
• Mappability Filtering: Repetitive regions of the genome with low uniqueness must be excluded to prevent multi-mapping artifacts from distorting coverage profiles.
• Copy Number Adjustment: Somatic copy number alterations in tumor-derived cfDNA can inflate or deflate local coverage, requiring computational adjustment to isolate the nucleosome positioning signal from the copy number signal.

The comprehensive study of cell-free DNA fragmentation patterns, including fragment length, end motifs, jagged ends, and nucleosome positioning. Fragmentomic features provide orthogonal information to somatic mutations, enabling tissue-of-origin deconvolution and epigenetic state inference even when driver mutations are absent. Key metrics include the windowed protection score (WPS) and orientation-aware fragment end analysis.

A quantitative metric that measures the depth of cfDNA coverage within a sliding window centered on a genomic position, normalized against the surrounding region. A high WPS indicates strong nucleosome occupancy and protection from nuclease digestion, while a low WPS corresponds to linker regions between nucleosomes. WPS profiles are used to infer transcription factor binding sites and generate genome-wide chromatin accessibility maps from plasma DNA.

The characterization of nucleotide preferences at cfDNA fragment termini, reflecting the sequence specificity of the nucleases responsible for DNA cleavage during apoptosis and necrosis. Common motifs include CCCA and CCTG trinucleotides. The frequency distribution of end motifs varies by tissue type and nuclease activity, providing a cell-type-specific signature that complements nucleosome positioning for determining the origin of circulating DNA.

A computational method that uses reference methylation atlases or nucleosome footprint profiles to estimate the relative contributions of different tissues to a cfDNA mixture. By comparing the observed fragmentation or methylation pattern against tissue-specific signatures from liver, lung, hematopoietic cells, and other organs, algorithms can identify aberrant contributions indicative of tumor-derived DNA, even without prior knowledge of driver mutations.

The process of reconstructing regulatory element activity from cfDNA fragmentation data. Because nucleosomes protect DNA from cleavage, regions of open chromatin—such as active promoters and enhancers—exhibit distinct fragmentation signatures. Machine learning models trained on matched ATAC-seq or DNase-seq data can predict gene regulatory landscapes directly from plasma DNA, revealing which genes were active in the cells that released the cfDNA.

A high-resolution extension of nucleosome footprinting that identifies localized protection of DNA at specific transcription factor binding sites. When a transcription factor occupies its cognate motif, it physically shields the DNA from nuclease cleavage, producing a characteristic dip in fragmentation coverage flanked by accessible regions. This pattern enables inference of active transcriptional regulators in the tissue of origin from cfDNA sequencing data.