Hematopoietic cell-free DNA distinguishes cancer patients with low tumor fraction from healthy controls

About this Repository

This repository documents the work for my Bioinformatics MSc Project module at Birkbeck, University of London. This work was undertaken over the course of a semester in the Krauthammer Lab at the University of Zurich, where my primary supervisor was Zsolt Balázs, MD, PhD.

Project Abstract

Liquid biopsies of blood cell-free DNA (cfDNA) and circulating tumor DNA (ctDNA) enable minimally invasive cancer detection and monitoring. Various cfDNA and ctDNA analytic techniques readily detect cancer in samples when tumor fraction, the fraction of ctDNA to cfDNA in a sample, is high. However, cancer is often not detected in samples with low tumor fraction. Here, we show that hematopoietic cfDNA is a potential cancer biomarker in samples with low tumor fraction. In our prospective, radiotherapy cohort of 22 cases and 7 controls, we find that the hematopoietic DNAse I hypersensitivity site (DHS) signals are significantly different in cancer samples with low tumor fraction compared to control samples. These signals also explain most of the variance in our cohort. In addition, our reanalysis of data from a published cohort of 86 pediatric sarcoma samples with low tumor fraction reveals significant differences between control and case hematopoietic DHS signals at diagnosis, therapy, relapse and remission. Furthermore, signals of hematopoietic transcription factor binding sites (TFBS) of 41 out of 56 of our low tumor fraction samples are significantly different from controls. These findings demonstrate the value of non-tumor cfDNA in cancer detection.

Graphical Summary of Methods

Respository Contents

In reference to the graphical summary of methods above, the repository contents include:
(A/B) Information about the study cohort and samples;
(C) ichorCNA¹ and t-MAD² copy number aberration (CNA) plots;
(D) LIQUORICE³ DHS data output and plots, script for the PCA of dip depth z-scores, scripts for the reanalysis of a pediatric sarcoma cohort⁴;
(E) TFBS command-line tool and data output, script to split GTRD⁵ ChiP-seq data by TF;
(F) Scripts comparing fragment coverage and windowed protection score (WPS) at promoter regions.

References

1. Adalsteinsson, V. A. et al. Scalable whole-exome sequencing of cell-free DNA reveals high concordance with metastatic tumors. Nature Communications 8, 1–13 (2017).
   
2. Mouliere, F. et al. Enhanced detection of circulating tumor DNA by fragment size analysis. Science Translational Medicine 10, (2018).
   
3. Peneder, P., Bock, C. & Tomazou, E. M. LIQUORICE: detection of epigenetic signatures in liquid biopsies based on whole-genome sequencing data. Bioinformatics Advances 2, (2022).
   
4. Peneder, P. et al. Multimodal analysis of cell-free DNA whole-genome sequencing for pediatric cancers with low mutational burden. Nature Communications 12, (2021).
   
5. Kolmykov, S. et al. GTRD: an integrated view of transcription regulation. Nucleic Acids Research 49, D104–D111 (2020).