CoChat
Sign Up
Cited by 287Genomics and Phylogenetic StudiesAlgorithms and Data CompressionFractal and DNA sequence analysis

The DOI matches its Crossref record; 43 of the 43 references it cites were found in a registry or an index.

Effects of GC Bias in Next-Generation-Sequencing Data on De Novo Genome Assembly

Open access
Chen, Yen-Chun, Liu, Tsunglin, Yu, Chun-Hui, Chiang, Tzen-Yuh, Hwang, Chi-Chuan·2013·PLoS ONE
DOI 10.1371/journal.pone.0062856PMC PMC3639258
Open this paper in CoChat — free accountOpen PDF

Abstract

Next-generation-sequencing (NGS) has revolutionized the field of genome assembly because of its much higher data throughput and much lower cost compared with traditional Sanger sequencing. However, NGS poses new computational challenges to de novo genome assembly. Among the challenges, GC bias in NGS data is known to aggravate genome assembly. However, it is not clear to what extent GC bias affects genome assembly in general. In this work, we conduct a systematic analysis on the effects of GC bias on genome assembly. Our analyses reveal that GC bias only lowers assembly completeness when the degree of GC bias is above a threshold. At a strong GC bias, the assembly fragmentation due to GC bias can be explained by the low coverage of reads in the GC-poor or GC-rich regions of a genome. This effect is observed for all the assemblers under study. Increasing the total amount of NGS data thus rescues the assembly fragmentation because of GC bias. However, the amount of data needed for a full rescue depends on the distribution of GC contents. Both low and high coverage depths due to GC bias lower the accuracy of assembly. These pieces of information provide guidance toward a better de novo genome assembly in the presence of GC bias.

Overview

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

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

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

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0062856

Cites

43 of 43 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Scatter plots of GC content and read coverage of real Illumina data. The data sets are from S. aureus USA300 (A) and S. aureus MRSA252 (B) genomes. Read coverage is normalized to the mean value, which is represented by a horizontal dashed line. A vertical dashed line denotes the mean GC content. The data points are fitted by a straight line and the slope is defined as the degree of GC bias. The two cases represent a negative and positive GC bias, respectively.
Figure 1 Scatter plots of GC content and read coverage of real Illumina data. The data sets are from S. aureus USA300 (A) and S. aureus MRSA252 (B) genomes. Read coverage is normalized to the mean value, which is represented by a horizontal dashed line. A vertical dashed line denotes the mean GC content. The data points are fitted by a straight line and the slope is defined as the degree of GC bias. The two cases represent a negative and positive GC bias, respectively.CC BY · PubMed Central
Figure 2 Correlation between the degree of GC bias and two statistics of GC contents. No correlation can be observed between the degree of GC bias (y-axis) and either the mean GC content (A) or the standard deviation of GC contents (B).
Figure 2 Correlation between the degree of GC bias and two statistics of GC contents. No correlation can be observed between the degree of GC bias (y-axis) and either the mean GC content (A) or the standard deviation of GC contents (B).CC BY · PubMed Central
Figure 3 Scatter plots of GC content and read coverage of the simulated data. From the S. aureus USA300 genome, we simulated reads of 50X coverage at three degrees of GC bias: negative slope −3.83 (A), slope zero (B), and positive slope 3.72 (C).
Figure 3 Scatter plots of GC content and read coverage of the simulated data. From the S. aureus USA300 genome, we simulated reads of 50X coverage at three degrees of GC bias: negative slope −3.83 (A), slope zero (B), and positive slope 3.72 (C).CC BY · PubMed Central
Figure 4 Completeness of assemblies of three bacterial genomes by eight assemblers, each treating nine data sets. The nine data sets at various degrees of GC bias (shown in different colors) are simulated from the genomes of three bacteria: E. coli (A), S. aureus (B), and M. tuberculosis (C). Assembly completeness is measured by the N50 length of the contigs after error corrections. The left and right columns show the results of assembly using simulated data of a 50X and 100X coverage, respectively. Note that the Velvet-SC assembly of the genome M. tuberculosis failed without a clear reason. At a 50X coverage, strong GC bias leads to more fragmented assembly in all cases. Such performance drops can be rescued via increasing the amount of data to a 100X coverage.
Figure 4 Completeness of assemblies of three bacterial genomes by eight assemblers, each treating nine data sets. The nine data sets at various degrees of GC bias (shown in different colors) are simulated from the genomes of three bacteria: E. coli (A), S. aureus (B), and M. tuberculosis (C). Assembly completeness is measured by the N50 length of the contigs after error corrections. The left and right columns show the results of assembly using simulated data of a 50X and 100X coverage, respectively. Note that the Velvet-SC assembly of the genome M. tuberculosis failed without a clear reason. At a 50X coverage, strong GC bias leads to more fragmented assembly in all cases. Such performance drops can be rescued via increasing the amount of data to a 100X coverage.CC BY · PubMed Central
Figure 5 Completeness of the E.coli assemblies using data of various coverage. Assembly completeness is measured by N50 length of the corrected contigs, which are output by eight assemblers when treating simulated reads of various coverage (50X, 100X, 250X, 500X, 1000X, and 2000X) at a zero (blue line) and a strong positive GC bias (slope 3.6, pink line).
Figure 5 Completeness of the E.coli assemblies using data of various coverage. Assembly completeness is measured by N50 length of the corrected contigs, which are output by eight assemblers when treating simulated reads of various coverage (50X, 100X, 250X, 500X, 1000X, and 2000X) at a zero (blue line) and a strong positive GC bias (slope 3.6, pink line).CC BY · PubMed Central
Figure 6 Percentage of unaligned reference sequences. The results are from the assemblies of three bacterial genomes: E. coli (A), S. aureus (B), and M. tuberculosis (C). Each of the eight assemblers treats data at a strong negative, zero, and strong positive GC bias.
Figure 6 Percentage of unaligned reference sequences. The results are from the assemblies of three bacterial genomes: E. coli (A), S. aureus (B), and M. tuberculosis (C). Each of the eight assemblers treats data at a strong negative, zero, and strong positive GC bias.CC BY · PubMed Central
Figure 7 Read coverage and mis-assemblies on the S.aureus genome. Read coverage (blue curves) and mis-assemblies (colored regions in the bottom bar) in a region of S. aureus genome at a strong negative (A), zero (B), and strong positive (C) GC bias. Different colors represent different types of mis-assemblies: tandem repeat error (green), translocation error (purple), unaligned reference sequence (red). The colors are projected to the curve of read coverage. The down-triangles in the coverage curves denote single-base insertions.
Figure 7 Read coverage and mis-assemblies on the S.aureus genome. Read coverage (blue curves) and mis-assemblies (colored regions in the bottom bar) in a region of S. aureus genome at a strong negative (A), zero (B), and strong positive (C) GC bias. Different colors represent different types of mis-assemblies: tandem repeat error (green), translocation error (purple), unaligned reference sequence (red). The colors are projected to the curve of read coverage. The down-triangles in the coverage curves denote single-base insertions.CC BY · PubMed Central
Figure 8 Distributions of coverage depths at all bases and at error bases. Distributions of coverage depths at error bases (red curves) are compared with those at all bases (blue curves) in the Velvet assemblies of three bacterial genomes: E. coli (A), S. aureus (B), and M. tuberculosis (C), using data simulated at a strong negative (left column), zero (middle column), and strong positive (right column) GC bias.
Figure 8 Distributions of coverage depths at all bases and at error bases. Distributions of coverage depths at error bases (red curves) are compared with those at all bases (blue curves) in the Velvet assemblies of three bacterial genomes: E. coli (A), S. aureus (B), and M. tuberculosis (C), using data simulated at a strong negative (left column), zero (middle column), and strong positive (right column) GC bias.CC BY · PubMed Central
Figure 9 Ratio of corrected N50 length at a strong GC bias to that at no GC bias. Ratio of the corrected N50 length at a strong negative GC bias (A) and a strong positive GC bias (B) to that at no GC bias when assembling the data of five species (in different colors) using eight assemblers.
Figure 9 Ratio of corrected N50 length at a strong GC bias to that at no GC bias. Ratio of the corrected N50 length at a strong negative GC bias (A) and a strong positive GC bias (B) to that at no GC bias when assembling the data of five species (in different colors) using eight assemblers.CC BY · PubMed Central
Figure 10 Distribution of GC contents and read coverage of the five species under study. The red curves stand for GC contents (scale in top axis). The blue and yellow curves represent read coverage at a strong positive and strong negative GC bias, respectively (scale in bottom axis). We used the data at 100X coverage for the five species.
Figure 10 Distribution of GC contents and read coverage of the five species under study. The red curves stand for GC contents (scale in top axis). The blue and yellow curves represent read coverage at a strong positive and strong negative GC bias, respectively (scale in bottom axis). We used the data at 100X coverage for the five species.CC BY · PubMed Central
Figure 11 Scatter plots of GC content and read coverage of data simulated with various degrees of background fluctuations. The data are simulated from the E. coli genome at three degrees of background fluctuations: zero (top row), 10 (middle row), and 20 (bottom row). At each degree of background fluctuation, we simulated PE reads at a strong negative (A), zero (B), and a strong positive (C) GC bias, respectively.
Figure 11 Scatter plots of GC content and read coverage of data simulated with various degrees of background fluctuations. The data are simulated from the E. coli genome at three degrees of background fluctuations: zero (top row), 10 (middle row), and 20 (bottom row). At each degree of background fluctuation, we simulated PE reads at a strong negative (A), zero (B), and a strong positive (C) GC bias, respectively.CC BY · PubMed Central
Figure 12 Corrected N50 length of assemblies at three background fluctuations. We show the corrected N50 length in eight assemblies of three bacterial genomes: E. coli (A), S. aureus (B), and M. tuberculosis (C), using simulated data at three degrees of background fluctuations (x-axis), each at three degrees of GC bias: negative (yellow), zero (dark blue), and positive (pink).
Figure 12 Corrected N50 length of assemblies at three background fluctuations. We show the corrected N50 length in eight assemblies of three bacterial genomes: E. coli (A), S. aureus (B), and M. tuberculosis (C), using simulated data at three degrees of background fluctuations (x-axis), each at three degrees of GC bias: negative (yellow), zero (dark blue), and positive (pink).CC BY · PubMed Central
Figure 13 Estimation of the degree of GC bias using reference sequences and assembled contigs. We show the scatter plots of GC content and read coverage for P.fluorescens Pf0-1 Illumina library (DRR001171) based on the known reference genome (A) and the contigs assembled by Edena, which contain 6,610,650 bases and the N50 length is 8,257 (B).
Figure 13 Estimation of the degree of GC bias using reference sequences and assembled contigs. We show the scatter plots of GC content and read coverage for P.fluorescens Pf0-1 Illumina library (DRR001171) based on the known reference genome (A) and the contigs assembled by Edena, which contain 6,610,650 bases and the N50 length is 8,257 (B).CC BY · PubMed Central
Figure 14 Correlation between the degree of GC bias obtained using reference sequences and assembled contigs. The correlation is calculated for thirteen Illumina data sets, including eight data sets by Edena, four data sets by Vevlet and one data set by ABySS. The high R2 value (0.88) indicates that estimating the degree of GC bias using the assembled contigs is appropriate.
Figure 14 Correlation between the degree of GC bias obtained using reference sequences and assembled contigs. The correlation is calculated for thirteen Illumina data sets, including eight data sets by Edena, four data sets by Vevlet and one data set by ABySS. The high R2 value (0.88) indicates that estimating the degree of GC bias using the assembled contigs is appropriate.CC BY · PubMed Central

Verification

Claims the paper makes

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

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

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

Show all 43 references

Cited by

Sorted by citations

See all 287

Related

Metadata

Authors
Chen, Y.-C., Liu, T., Yu, C.-H., Chiang, T.-Y., & Hwang, C.-C.
Affiliations
National Cheng Kung University (TW)
Corresponding author
Hwang, Chi-Chuan
ORCID
Venue
PLoS ONE
Publisher
Public Library of Science (PLoS)
Published
2013-04-29
Volume
8
Issue
4
Pages
e62856
ISSN
1932-6203 (ISSN-L)
DOI
10.1371/journal.pone.0062856
PMC
PMC3639258
OpenAlex
W1998330439
Cited by
285
References
43
Citations by year
2026: 7 · 2025: 22 · 2024: 25 · 2023: 20 · 2022: 26 · 2021: 48 · 2020: 30 · 2019: 13 · 2018: 27 · 2017: 23 · 2016: 18 · 2015: 11 · 2014: 12 · 2013: 2
FWCI
6.42
Type
journal article
Language
en
Open access
Licence
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, other-oa)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Figshare (submittedVersion, cc-by)
Funders
National Science Council
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
  • Physical Sciences › Mathematics › Theoretical Computer Science › Algorithms and Data Compression
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Fractal and DNA sequence analysis
Keywords
GC content bias, de novo genome assembly, next-generation sequencing, assembly completeness, read coverage, assembly accuracy
MeSH terms
Reference Standards, Artifacts, Genome, Bacterial, GC Rich Sequence, Genomics, High-Throughput Nucleotide Sequencing

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