Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 54 of the 54 references it cites were found in a registry or an index; 17 of the 20 sources its claims cite were checked for support.

Removing reference bias and improving indel calling in ancient DNA data analysis by mapping to a sequence variation graph

Open access

Published version of a preprint

Abstract

Abstract Background During the last decade, the analysis of ancient DNA (aDNA) sequence has become a powerful tool for the study of past human populations. However, the degraded nature of aDNA means that aDNA molecules are short and frequently mutated by post-mortem chemical modifications. These features decrease read mapping accuracy and increase reference bias, in which reads containing non-reference alleles are less likely to be mapped than those containing reference alleles. Alternative approaches have been developed to replace the linear reference with a variation graph which includes known alternative variants at each genetic locus. Here, we evaluate the use of variation graph software to avoid reference bias for aDNA and compare with existing methods. Results We use to align simulated and real aDNA samples to a variation graph containing 1000 Genome Project variants and compare with the same data aligned with to the human linear reference genome. Using leads to a balanced allelic representation at polymorphic sites, effectively removing reference bias, and more sensitive variant detection in comparison with , especially for insertions and deletions (indels). Alternative approaches that use relaxed parameter settings or filter alignments can also reduce bias but can have lower sensitivity than , particularly for indels. Conclusions Our findings demonstrate that aligning aDNA sequences to variation graphs effectively mitigates the impact of reference bias when analyzing aDNA, while retaining mapping sensitivity and allowing detection of variation, in particular indel variation, that was previously missed.

Overview

CoChat Summary

TL;DR

The paper shows that mapping ancient DNA reads to a variation graph with `vg` largely removes reference bias and improves indel detection compared with standard `bwa aln` mapping. On both simulated and 34 real ancient human samples, `vg` gives more balanced allele representation and higher sensitivity, though at a cost of longer runtime and greater memory use.

Key findings

  • In simulation, `bwa aln -n 0.02` produced an average alternate-allele fraction of 0.48267 (95% CI [0.48095, 0.48438]), while `vg` gave 0.49988 (95% CI [0.49984, 0.49991]), essentially eliminating reference bias.
  • On the high-coverage Yamnaya sample, `vg` recovered an alternate-allele read fraction of 0.4925 (95% CI [0.4914, 0.4937]) versus 0.4742 for `bwa aln -n 0.02` and 0.4773 for `bwa aln -n 0.01 -o 2`.
  • At 4× coverage, `vg` recovered about 13% more heterozygotes than `bwa aln` in the downsampling experiment; for indels, `vg` recovered approximately twice as many as `bwa aln` at high coverage and even more at lower coverage.
  • `bwa aln` failed to detect indels longer than 7 bp, whereas `vg` recovered indels up to at least 18 bp in the PCA/indel analysis; the CCR5 delta 32 deletion was detected in the 4900-year-old Yamnaya sample and three more recent ancient British samples.

Methods

  • Compared `vg` mapping to a 1000 Genomes Project variation graph against `bwa aln` mapping to the GRCh37/hs37d5 linear reference using both simulated reads and 34 previously published ancient human samples spanning multiple archaeological contexts and sequencing strategies.
  • Simulations generated 50-bp reads across chromosome 11 Human Origins SNP sites, added alternate alleles to half the reads, and introduced post-mortem deamination based on damage profiles from 102 ancient genomes; additional microbial-read simulations tested false alignments.
  • Main analyses included allele-balance and mapping-accuracy comparisons, D-statistics, PCA, variant calling with bcftools/FreeBayes, and downsampling of the high-coverage Yamnaya individual to test recovery at lower coverage.

Limitations

  • The paper states that `vg` alignment takes about 5× longer than bwa aln -n 0.01 -o 2 and 29× longer than bwa mem, and requires substantially more memory for indexing and mapping.
  • It notes that mapping quality scores are not directly comparable between `vg` and `bwa aln`, so comparisons depend on chosen thresholds.
  • It also states that the graph approach relies on a graph built from present-day human variation, so it is not appropriate for archaic populations such as Neanderthals and Denisovans; this is a limitation the paper explicitly discusses.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1186/s13059-020-02160-7

Cites

54 of 54 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1 Sequence tube maps. Sequence tube maps [19] of a small region of the human genome with aDNA reads from the Yamnaya individual aligned with abwa aln to a linear reference sequence and bvg map to a graph containing 1000 Genomes variants. The individual is heterozygous for both an indel (GTTTGAG/-) and a SNP (A/C) in this region, with insertion and alternate allele on the same haplotype. The two underlying haplotypes in this region are colored in gray, and red and blue lines indicate forward and reverse reads, respectively. None of the 6 reads across the insertion and only 2 of 12 reads across the SNP were mapped by bwa. Reads were locally realigned with vg map to the graph for the purpose of visualization
Fig. 1 Sequence tube maps. Sequence tube maps [19] of a small region of the human genome with aDNA reads from the Yamnaya individual aligned with abwa aln to a linear reference sequence and bvg map to a graph containing 1000 Genomes variants. The individual is heterozygous for both an indel (GTTTGAG/-) and a SNP (A/C) in this region, with insertion and alternate allele on the same haplotype. The two underlying haplotypes in this region are colored in gray, and red and blue lines indicate forward and reverse reads, respectively. None of the 6 reads across the insertion and only 2 of 12 reads across the SNP were mapped by bwa. Reads were locally realigned with vg map to the graph for the purpose of visualizationCC BY · PubMed Central
Fig. 2 Comparing vg graph, bwa aln, and bwa mem using simulated ancient DNA. Comparing bwa aln and vg map performance when aligning reads simulated from chromosome 11 of the Human Origins panel. Lines represent ordinary least squares (OLS) regression results for the allele/aligner conditions corresponding to their colors. a Comparison between vg graph and bwa aln -n 0.02. b Comparison between vg graph and bwa aln -n 0.01 -o 2. c Comparison of the mean percentage (and 95% CI) of mapped reads in simulated data by vg graph, bwa aln, and bwa mem using different alignment parameters and minimum mapping quality filtering thresholds. d Mean alternate allele fraction (and 95% CI) of simulated reads after alignment with the different methods and minimum mapping quality filtering thresholds. We also show results obtained after processing simulated data with two previously published workflows for addressing reference bias: modified reads (“modreads”) [10] and modified reference genome (“altref genome”) [15]
Fig. 2 Comparing vg graph, bwa aln, and bwa mem using simulated ancient DNA. Comparing bwa aln and vg map performance when aligning reads simulated from chromosome 11 of the Human Origins panel. Lines represent ordinary least squares (OLS) regression results for the allele/aligner conditions corresponding to their colors. a Comparison between vg graph and bwa aln -n 0.02. b Comparison between vg graph and bwa aln -n 0.01 -o 2. c Comparison of the mean percentage (and 95% CI) of mapped reads in simulated data by vg graph, bwa aln, and bwa mem using different alignment parameters and minimum mapping quality filtering thresholds. d Mean alternate allele fraction (and 95% CI) of simulated reads after alignment with the different methods and minimum mapping quality filtering thresholds. We also show results obtained after processing simulated data with two previously published workflows for addressing reference bias: modified reads (“modreads”) [10] and modified reference genome (“altref genome”) [15]CC BY · PubMed Central
Fig. 3 Downsampling a high-coverage aDNA sample. The comparative effect of downsampling on heterozygous variant calling following bwa aln and vg map alignment of reads from the ancient Yamnaya sample [26] with different parameters and mapping quality filtering thresholds, and including post-processing of bwa aln with the modified read filter [10]. a SNPs. b Indels (the modified read filter does not apply in this case)
Fig. 3 Downsampling a high-coverage aDNA sample. The comparative effect of downsampling on heterozygous variant calling following bwa aln and vg map alignment of reads from the ancient Yamnaya sample [26] with different parameters and mapping quality filtering thresholds, and including post-processing of bwa aln with the modified read filter [10]. a SNPs. b Indels (the modified read filter does not apply in this case)CC BY · PubMed Central
Fig. 4 Comparison between vg and bwa aln for indel detection. a Alternate allele observations at indels. b Comparison between vg graph and bwa aln in the detection of the CCR5 delta 32 deletion associated with HIV-1 resistance. Reads containing the deletion were mapped with vg in four ancient samples, but not with bwa
Fig. 4 Comparison between vg and bwa aln for indel detection. a Alternate allele observations at indels. b Comparison between vg graph and bwa aln in the detection of the CCR5 delta 32 deletion associated with HIV-1 resistance. Reads containing the deletion were mapped with vg in four ancient samples, but not with bwaCC BY · PubMed Central

Verification

Claims the paper makes

13 Valid · 0 Not valid · 1 Not checked · 59 Own results and interpretations

35 more claims · 3 with their cited sources checked

Create a free account to see all 73 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 4
  • It may state two separate claims as one. 29
  • The quote only partly states this claim. 1

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Citations

Cites

Show all 54 references

Research works printed but not linked: 1

  1. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G, Durbin R. The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009; 25(16):2078–9. · DOI · Paper page

Cited by

See all 128

Metadata

Authors
Martiniano, R., Garrison, E., Jones, E. R., Manica, A., & Durbin, R.
Affiliations
  • University of Cambridge (GB)
  • University of California, Santa Cruz (US)
Corresponding author
Durbin, Richard
Venue
Genome Biology
Published
2020-09-17
Volume
21
Issue
1
Pages
250-250
ISSN
  • 1465-6906 (ISSN-L)
  • 1474-760X
OpenAlex
W3086246013
Cited by
128
References
54
Citations by year
2026: 10 · 2025: 56 · 2024: 20 · 2023: 15 · 2022: 17 · 2021: 8 · 2020: 2
FWCI
3.40
Type
journal article
Language
en
Open access
Licence
Locations
  • Genome biology (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Liverpool John Moores University (acceptedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Apollo (University of Cambridge) (submittedVersion, cc-by)
  • Apollo (cc-by)
  • Apollo
Funders
Wellcome Trust (WT207492, 207492/Z/17/Z, WT206194)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Forensic and Genetic Research
  • Physical Sciences › Mathematics › Theoretical Computer Science › Algorithms and Data Compression
Keywords
ancient DNA, reference bias, variation graphs, indel calling, sequence alignment, VG, BWA, variant calling
MeSH terms
DNA, Ancient, Humans, Reference Standards, Genome, Human, Sequence Analysis, DNA, INDEL Mutation

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Removing reference bias and improving indel calling in ancient DNA data analysis by mapping to a sequence variation graph. https://app.cochat.ai/papers/10.1186/s13059-020-02160-7
The paper
Martiniano, R., Garrison, E., Jones, E. R., Manica, A., & Durbin, R. (2020). Removing reference bias and improving indel calling in ancient DNA data analysis by mapping to a sequence variation graph. Genome Biology. https://doi.org/10.1186/s13059-020-02160-7

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.