Back

Mapping Transcription Factor-Nucleosome Dynamics from Plasma cfDNA

Rao, S.; Han, A. L.; Zukowski, A.; Kopin, E.; Sartorius, C. A.; Kabos, P.; Ramachandran, S.

2021-04-15 genomics
10.1101/2021.04.14.439883 bioRxiv
Show abstract

Cell-free DNA (cfDNA) contains a composite map of the epigenomes of its cells-of-origin. Tissue-specific transcription factor (TF) binding inferred from cfDNA could enable us to track disease states in humans in a minimally invasive manner. Here, by enriching for short cfDNA fragments, we directly map TF footprints at single binding sites from plasma. We show that the enrichment of TF footprints in plasma reflects the binding strength of the TF in cfDNA tissue-of-origin. Based on this principle, we were able to identify the subset of genome-wide binding sites for selected TFs that leave TF-specific footprints in plasma. These footprints enabled us to not only identify the tissue-of-origin of cfDNA but also map the chromatin structure around the factor-bound sites in their cells-of-origin. To ask if we can use these plasma TF footprints to map cancer states, we first defined pure tumor TF signatures in plasma in vivo using estrogen receptor-positive (ER+) breast cancer xenografts. We found that the tumor-specific cfDNA protections of ER- could distinguish WT, ER-amplified, and ER-mutated xenografts. Further, tumor-specific cfDNA protections of ER- and FOXA1 reflect TF-specific accessibility across human breast tumors, demonstrating our ability to capture tumor TF binding in plasma. We then scored TF binding in human plasma samples and identified specific binding sites whose plasma TF protections can identify the presence of cancer and specifically breast cancer. Thus, plasma TF footprints enable minimally invasive mapping of the regulatory landscape of cancer in humans.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.