Revisiting the Briggs ancient DNA damage model: a fast regression method to estimate postmortem damage
Preprint
Abstract
Motivation One essential initial step in the analysis of ancient DNA is to authenticate its ancientness to ensure reliable conclusions. That is, meticulously assessing whether next-generation sequencing reads exhibit ancient characteristics, with a particular focus on the postmortem damage (PMD) signal induced by cytosine deamination in the fragments termini. We present a novel statistical method implementation in a fast multithreaded program ngsBriggs that enables the rapid quantification of PMD by calculation of the Briggs ancient damage model parameters (Briggs parameters). Results Using a fast multinomial regression approach, ngsBriggs accurately models the Briggs parameters, quantifying the PMD signal from single and double-stranded DNA regions. We revisit and extend the original Briggs model, with ngsBriggs modeling PMD signals for contemporary sequencing platforms. Furthermore, ngsBriggs asserts itself as a reliable and consistent tool, by accurately estimating the Briggs parameters across a variety of contamination levels. The classification accuracy of ngsBriggs significantly exceeds the current tool available when discerning ancient-from modern sequencing reads to decontaminate samples. Our novel method and implementation ngsBriggs outperforms existing tools regarding computational speed and accuracy, establishing its practicality and usability. Our tool, ngsBriggs offers a practical and accurate toolset for researchers seeking to authenticate ancient DNA and improve the quality of their data. Availability https://github.com/lz398/metadamage_briggsKeywords
This publication has 20 references indexed in Scilit:
- NGSNGS: next-generation simulator for next-generation sequencing dataBioinformatics, 2023
- A 2-million-year-old ecosystem in Greenland uncovered by environmental DNANature, 2022
- Population genomics of the Viking worldNature, 2020
- Population genomics of Bronze Age EurasiaNature, 2015
- Separating endogenous ancient DNA from modern day contamination in a Siberian NeandertalProceedings of the National Academy of Sciences, 2014
- mapDamage2.0: fast approximate Bayesian estimates of ancient DNA damage parametersBioinformatics, 2013
- A Revised Timescale for Human Evolution Based on Ancient Mitochondrial GenomesCurrent Biology, 2013
- mapDamage: testing for damage patterns in ancient DNA sequencesBioinformatics, 2011
- The Sequence Alignment/Map format and SAMtoolsBioinformatics, 2009
- Patterns of damage in genomic DNA sequences from a NeandertalProceedings of the National Academy of Sciences of the United States of America, 2007