CoChat

Substantial biases in ultra-short read data sets from high-throughput DNA sequencing

Open access

The DOI matches its Crossref record; 22 of the 22 references it cites were found in a registry or an index.

Abstract

Novel sequencing technologies permit the rapid production of large sequence data sets. These technologies are likely to revolutionize genetics and biomedical research, but a thorough characterization of the ultra-short read output is necessary. We generated and analyzed two Illumina 1G ultra-short read data sets, i.e. 2.8 million 27mer reads from a Beta vulgaris genomic clone and 12.3 million 36mers from the Helicobacter acinonychis genome. We found that error rates range from 0.3% at the beginning of reads to 3.8% at the end of reads. Wrong base calls are frequently preceded by base G. Base substitution error frequencies vary by 10- to 11-fold, with A > C transversion being among the most frequent and C > G transversions among the least frequent substitution errors. Insertions and deletions of single bases occur at very low rates. When simulating re-sequencing we found a 20-fold sequencing coverage to be sufficient to compensate errors by correct reads. The read coverage of the sequenced regions is biased; the highest read density was found in intervals with elevated GC content. High Solexa quality scores are over-optimistic and low scores underestimate the data quality. Our results show different types of biases and ways to detect them. Such biases have implications on the use and interpretation of Solexa data, for de novo sequencing, re-sequencing, the identification of single nucleotide polymorphisms and DNA methylation sites, as well as for transcriptome analysis.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1093/nar/gkn425

Cites

22 of 22 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1. Pie charts of the read analysis with ELAND. The ELAND categories are: QC: no matching done because of low quality of the read (more than two positions with quality score = −5), NM, no match found; U0, unique exact match found; U1, unique match with one error; U2, unique match with two errors; R0, multiple exact matches found; R1, multiple matches with one error; R2, multiple matches with two errors. The categories R0, R1, R2 are shown as a single entity. (a) ELAND categorizations for 27mer reads from Beta vulgaris clone ZR-47B15 (2 788 286 in total). (b) ELAND categorizations for 32mer reads from Helicobacter acinonychis (12 288 791 in total, trimmed by the last four base calls of the original 36mer data).
Figure 1. Pie charts of the read analysis with ELAND. The ELAND categories are: QC: no matching done because of low quality of the read (more than two positions with quality score = −5), NM, no match found; U0, unique exact match found; U1, unique match with one error; U2, unique match with two errors; R0, multiple exact matches found; R1, multiple matches with one error; R2, multiple matches with two errors. The categories R0, R1, R2 are shown as a single entity. (a) ELAND categorizations for 27mer reads from Beta vulgaris clone ZR-47B15 (2 788 286 in total). (b) ELAND categorizations for 32mer reads from Helicobacter acinonychis (12 288 791 in total, trimmed by the last four base calls of the original 36mer data).CC BY-NC · PubMed Central
Figure 2. Correlation of the Solexa read coverage and GC content. (a) 27mer reads generated from Beta vulgaris BAC ZR-47B15 (b) 32mer data set from the Helicobacter acinonychis genome. Each data point corresponds to the number of reads recorded for a 1-kbp window (shift of 100 bp in Beta and 1 kbp in Helicobacter).
Figure 2. Correlation of the Solexa read coverage and GC content. (a) 27mer reads generated from Beta vulgaris BAC ZR-47B15 (b) 32mer data set from the Helicobacter acinonychis genome. Each data point corresponds to the number of reads recorded for a 1-kbp window (shift of 100 bp in Beta and 1 kbp in Helicobacter).CC BY-NC · PubMed Central
Figure 3. Distribution of Solexa reads along the reference sequences considering unique match positions reported by ELAND (zero, one or two mismatch bases) and reads with more than one match position (no mismatch bases) detected with a Perl script. (a) Read distribution along the Beta vulgaris BAC sequence (with cloning vector pBeloBACII). 2 166 892 27mer reads were matched against the finished sequence (enclosed by the cloning vector,∼117 kbp in total). The read coverage was calculated in 200 consecutive 0.58 kbp windows. (b) Read distribution along the 1.55 Mbp Helicobacter genome, based on 8 700 113 32mer reads. The local coverage is shown in 200 consecutive windows of 7.77 kbp.
Figure 3. Distribution of Solexa reads along the reference sequences considering unique match positions reported by ELAND (zero, one or two mismatch bases) and reads with more than one match position (no mismatch bases) detected with a Perl script. (a) Read distribution along the Beta vulgaris BAC sequence (with cloning vector pBeloBACII). 2 166 892 27mer reads were matched against the finished sequence (enclosed by the cloning vector,∼117 kbp in total). The read coverage was calculated in 200 consecutive 0.58 kbp windows. (b) Read distribution along the 1.55 Mbp Helicobacter genome, based on 8 700 113 32mer reads. The local coverage is shown in 200 consecutive windows of 7.77 kbp.CC BY-NC · PubMed Central
Figure 4. Frequency of wrong base calls in Solexa reads depending on the position along the read (27mer reads from Beta vulgaris and 32mer reads from Helicobacter). (a) Error frequency per position calculated from considering wrong base calls only. The highest error frequency is observed at the read 3′ end. (b) Per-base error rates (overall error frequency per position considering all base calls).
Figure 4. Frequency of wrong base calls in Solexa reads depending on the position along the read (27mer reads from Beta vulgaris and 32mer reads from Helicobacter). (a) Error frequency per position calculated from considering wrong base calls only. The highest error frequency is observed at the read 3′ end. (b) Per-base error rates (overall error frequency per position considering all base calls).CC BY-NC · PubMed Central
Figure 5. Compensation of sequencing errors by deep sequencing in re-sequencing projects. The average number of errors per kbp is shown for different levels of coverage. For coverages below 2, reads are unlikely to overlap and compensation of sequencing errors is rare (thus, sequencing errors accumulate when the coverage is increased). For coverages above 3-fold the number of uncompensated errors drops rapidly with the increase of coverage.
Figure 5. Compensation of sequencing errors by deep sequencing in re-sequencing projects. The average number of errors per kbp is shown for different levels of coverage. For coverages below 2, reads are unlikely to overlap and compensation of sequencing errors is rare (thus, sequencing errors accumulate when the coverage is increased). For coverages above 3-fold the number of uncompensated errors drops rapidly with the increase of coverage.CC BY-NC · PubMed Central
Figure 6. Distance between two errors on a read in the Helicobacter and Beta vulgaris data sets. ‘0’ indicates that the erroneous base calls are next to each other.
Figure 6. Distance between two errors on a read in the Helicobacter and Beta vulgaris data sets. ‘0’ indicates that the erroneous base calls are next to each other.CC BY-NC · PubMed Central
Figure 7. Sequence context of wrong base calls in Solexa reads from Helicobacter acinonychis and Beta vulgaris, considering one base upstream and downstream of the wrong base calls. An ‘e’ indicates the substituted base. The scatterplot shows the correlation of the relative frequencies (relating the frequency of 3-tuples at error positions to the frequency of all 3-tuples in the reads) for the two data sets.
Figure 7. Sequence context of wrong base calls in Solexa reads from Helicobacter acinonychis and Beta vulgaris, considering one base upstream and downstream of the wrong base calls. An ‘e’ indicates the substituted base. The scatterplot shows the correlation of the relative frequencies (relating the frequency of 3-tuples at error positions to the frequency of all 3-tuples in the reads) for the two data sets.CC BY-NC · PubMed Central
Figure 8. Frequency of substitution errors in the Helicobacter acinonychis and Beta vulgaris Solexa read data sets.
Figure 8. Frequency of substitution errors in the Helicobacter acinonychis and Beta vulgaris Solexa read data sets.CC BY-NC · PubMed Central
Figure 9. Histograms of base quality values for all correct base calls (a) and all wrong base calls (b) in the Beta and Helicobacter data sets.
Figure 9. Histograms of base quality values for all correct base calls (a) and all wrong base calls (b) in the Beta and Helicobacter data sets.CC BY-NC · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

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?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Cited by

See all 1,088

Metadata

Authors
Dohm, J. C., Lottaz, C., Borodina, T., & Himmelbauer, H.
Affiliations
  • Max Planck Institute for Molecular Genetics (DE)
  • University of Regensburg (DE)
  • Max Planck Society (DE)
Corresponding author
Himmelbauer, Heinz
Venue
Nucleic Acids Research
Published
2008-07-26
Volume
36
Issue
16
Pages
e105-e105
ISSN
  • 0305-1048 (ISSN-L)
  • 1362-4962
OpenAlex
W2131088968
Cited by
1,088
References
22
Citations by year
2026: 14 · 2025: 18 · 2024: 17 · 2023: 20 · 2022: 36 · 2021: 40 · 2020: 46 · 2019: 52 · 2018: 45 · 2017: 70 · 2016: 69 · 2015: 76 · 2014: 104 · 2013: 109 · 2012: 100
FWCI
37.26
Type
journal article
Language
en
Open access
Locations
  • Nucleic Acids Research (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • University of Regensburg Publication Server (University of Regensburg) (acceptedVersion)
  • Max Planck Institute for Plasma Physics (submittedVersion)
  • University of Regensburg Publication Server (University of Regensburg)
Funders
Max-Planck-Gesellschaft
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Fractal and DNA sequence analysis
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene expression and cancer classification
Keywords
sequencing error rate, insertion and deletion errors, GC content, resequencing, Beta vulgaris, de novo sequencing, single nucleotide polymorphisms, gene expression profiling
MeSH terms
Base Composition, DNA, Software, Helicobacter, Sequence Deletion, Sequence Analysis, DNA, Beta vulgaris

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.

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

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.