CoChat

Characterization of fusion genes and the significantly expressed fusion isoforms in breast cancer by hybrid sequencing

Open access

The DOI matches its Crossref record; 56 of the 60 references it cites were found in a registry or an index.

Abstract

We developed an innovative hybrid sequencing approach, IDP-fusion, to detect fusion genes, determine fusion sites and identify and quantify fusion isoforms. IDP-fusion is the first method to study gene fusion events by integrating Third Generation Sequencing long reads and Second Generation Sequencing short reads. We applied IDP-fusion to PacBio data and Illumina data from the MCF-7 breast cancer cells. Compared with the existing tools, IDP-fusion detects fusion genes at higher precision and a very low false positive rate. The results show that IDP-fusion will be useful for unraveling the complexity of multiple fusion splices and fusion isoforms within tumorigenesis-relevant fusion genes.

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/gkv562

Cites

56 of 56 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1. Flowchart of IDP-fusion. IDP-fusion contains three main steps: (1) detect fusion genes by genome-wide alignments of long reads (e.g. using GMAP); (2) determine precise fusion sites by alignment of short reads to Artificial Reference Sequences (ARSs) (e.g. using STAR); (3) identify and quantify significantly expressed isoforms, including fusion isoforms, from fusion genes.
Figure 1. Flowchart of IDP-fusion. IDP-fusion contains three main steps: (1) detect fusion genes by genome-wide alignments of long reads (e.g. using GMAP); (2) determine precise fusion sites by alignment of short reads to Artificial Reference Sequences (ARSs) (e.g. using STAR); (3) identify and quantify significantly expressed isoforms, including fusion isoforms, from fusion genes.CC BY · PubMed Central
Figure 2. Isoform candidate constructions from fusion genes. Isoform candidate constructions contain three structural units: 5′/3′ ends, non-redundant splice linkages (including fusion splice linkages) and splices. All possible splice combinations are included until reaching the 5′ end and 3′ end. Both regular isoforms and fusion isoforms are constructed.
Figure 2. Isoform candidate constructions from fusion genes. Isoform candidate constructions contain three structural units: 5′/3′ ends, non-redundant splice linkages (including fusion splice linkages) and splices. All possible splice combinations are included until reaching the 5′ end and 3′ end. Both regular isoforms and fusion isoforms are constructed.CC BY · PubMed Central
Figure 3. Precisions of fusion gene detections by IDP-fusion, six SGS-only methods and a TGS-only method. (A) The total numbers of fusion genes detected from MCF-7 cells by IDP-fusion, six SGS-only methods and Iso-Seq (TGS-only method) and the corresponding numbers of the gold standard fusion genes are shown by stacked bars. The precisions are also shown as the rates of the gold standard fusion genes detected. Two modes of IDP-fusion, using either ≥2 supporting long reads (LR) and ≥1 supporting short read (SR) or ≥2 supporting long reads and ≥2 supporting short reads (default settings) are compared with six SGS-only methods that were run under their default settings and Iso-Seq method requiring ≥5 full-length long reads. (B) The precision of fusion gene detection by IDP-fusion is higher than SGS-only methods, regardless of increase of the minimum numbers of supporting short reads. The precision of the default setting for each method is labeled by ‘x’. Note that BreakFusion is not shown because the software does not output the number of supporting reads for each fusion site.
Figure 3. Precisions of fusion gene detections by IDP-fusion, six SGS-only methods and a TGS-only method. (A) The total numbers of fusion genes detected from MCF-7 cells by IDP-fusion, six SGS-only methods and Iso-Seq (TGS-only method) and the corresponding numbers of the gold standard fusion genes are shown by stacked bars. The precisions are also shown as the rates of the gold standard fusion genes detected. Two modes of IDP-fusion, using either ≥2 supporting long reads (LR) and ≥1 supporting short read (SR) or ≥2 supporting long reads and ≥2 supporting short reads (default settings) are compared with six SGS-only methods that were run under their default settings and Iso-Seq method requiring ≥5 full-length long reads. (B) The precision of fusion gene detection by IDP-fusion is higher than SGS-only methods, regardless of increase of the minimum numbers of supporting short reads. The precision of the default setting for each method is labeled by ‘x’. Note that BreakFusion is not shown because the software does not output the number of supporting reads for each fusion site.CC BY · PubMed Central
Figure 4. Precise fusion site determination of the fusion gene AIB1–chr1:107078407. (A) A fusion between AIB1 and an unannotated region of chromosome 1 was detected by six long reads (blue blocks), but the long read alignment ends are not in agreement on the precise fusion site. The alignment ends of the long read fragments span 37 and 65 bp at two sides, respectively. IDP-fusion aligned 31 short reads (black blocks) to the fusion site precisely at chr20:46130763–chr1:107078407. In particular, chr20:46130763 is the 3′ end of the first exon of AIB1 and the canonical splicing signal is found in this fusion site. (B) The fusion site chr20:46130763–chr1:107078407 was PCR validated from MCF-7 cDNA, but not healthy breast cDNA (BC) or a genomic control (GC).
Figure 4. Precise fusion site determination of the fusion gene AIB1–chr1:107078407. (A) A fusion between AIB1 and an unannotated region of chromosome 1 was detected by six long reads (blue blocks), but the long read alignment ends are not in agreement on the precise fusion site. The alignment ends of the long read fragments span 37 and 65 bp at two sides, respectively. IDP-fusion aligned 31 short reads (black blocks) to the fusion site precisely at chr20:46130763–chr1:107078407. In particular, chr20:46130763 is the 3′ end of the first exon of AIB1 and the canonical splicing signal is found in this fusion site. (B) The fusion site chr20:46130763–chr1:107078407 was PCR validated from MCF-7 cDNA, but not healthy breast cDNA (BC) or a genomic control (GC).CC BY · PubMed Central
Figure 5. The numbers of fusion sites determined from MCF-7 cells and four normal samples. Compared with the breast cancer cells MCF-7, four negative controls (human embryonic stem cell line/hESC, and human brain, liver and heart) are expected to have negligible gene fusion events. As increasing the requirement of supporting long reads, the numbers of fusion sites decreases. The dramatic drop occurs at the requirement of ≥2 supporting long reads. The fusion sites determined from the negative controls are negligible.
Figure 5. The numbers of fusion sites determined from MCF-7 cells and four normal samples. Compared with the breast cancer cells MCF-7, four negative controls (human embryonic stem cell line/hESC, and human brain, liver and heart) are expected to have negligible gene fusion events. As increasing the requirement of supporting long reads, the numbers of fusion sites decreases. The dramatic drop occurs at the requirement of ≥2 supporting long reads. The fusion sites determined from the negative controls are negligible.CC BY · PubMed Central
Figure 6. The distribution of fusion sites and fusion isoform counts in fusion genes of MCF-7 cells. (A) Single fusion sites were determined from 25 fusion genes and multiple fusion sites in the other 10 fusion genes. In particular, six fusion sites were determined in both fusion genes AIB1–chr1:107078407 and BCAS4–BCAS3. (B) Significantly expressed fusion isoforms (RPKM > 10) were identified from 14 fusion genes. Eight significantly expressed fusion isoforms were identified from BCAS4-BCAS3 (Figure 7A).
Figure 6. The distribution of fusion sites and fusion isoform counts in fusion genes of MCF-7 cells. (A) Single fusion sites were determined from 25 fusion genes and multiple fusion sites in the other 10 fusion genes. In particular, six fusion sites were determined in both fusion genes AIB1–chr1:107078407 and BCAS4–BCAS3. (B) Significantly expressed fusion isoforms (RPKM > 10) were identified from 14 fusion genes. Eight significantly expressed fusion isoforms were identified from BCAS4-BCAS3 (Figure 7A).CC BY · PubMed Central
Figure 7. Illustration of fusion isoforms occurring through alternative fusion splices or alternative regular splices. (A) Eight significantly expressed fusion isoforms were identified and quantified in the fusion gene BCAS4-BCAS3, sharing three fusion splices. Besides alternative fusion splices, the regular alternative splices within BCAS3 also contribute to the fusion isoform generation. (B) Three significantly expressed fusion isoforms were identified and quantified in the fusion gene RPS6KB1–VMP1, sharing two fusion splices. Besides alternative fusion splices, the regular alternative splices within both RPS6KB1 and VMP1 also contribute to the fusion isoform generation. (C) Two significantly expressed fusion isoforms were identified and quantified in the fusion gene ARFGEF2–SULF2, sharing only one fusion splice. The diversity of fusion isoform expression is driven by the regular alternative splicing of SULF2.
Figure 7. Illustration of fusion isoforms occurring through alternative fusion splices or alternative regular splices. (A) Eight significantly expressed fusion isoforms were identified and quantified in the fusion gene BCAS4-BCAS3, sharing three fusion splices. Besides alternative fusion splices, the regular alternative splices within BCAS3 also contribute to the fusion isoform generation. (B) Three significantly expressed fusion isoforms were identified and quantified in the fusion gene RPS6KB1–VMP1, sharing two fusion splices. Besides alternative fusion splices, the regular alternative splices within both RPS6KB1 and VMP1 also contribute to the fusion isoform generation. (C) Two significantly expressed fusion isoforms were identified and quantified in the fusion gene ARFGEF2–SULF2, sharing only one fusion splice. The diversity of fusion isoform expression is driven by the regular alternative splicing of SULF2.CC BY · PubMed Central
Figure 8. IDP-fusion detected and annotated fusion genes with novel gene and novel exons involved. (A) IDP-fusion detected the fusion gene between AIB1 and an unannotated region in chromosome 1. Eight fusion isoforms were estimated at the modest abundance (RPKM > 1) and contain seven novel exons (blue) annotated by IDP-fusion. (B) A selection of fusion long read alignments (bright blue) contributing structural information to isoform reconstruction is displayed in line with four reference annotations libraries: UCSC, RefSeq, GENCODE and Ensembl. The long read fragment alignments to AIB1 gene locus are shown at the left and the ones to the novel gene locus at the right. The long read alignments annotated a novel gene with seven exons not reported by the reference annotation libraries. (C) IDP-fusion detected the fusion gene UNK–ABCA5 and two corresponding fusion isoforms with RPKM > 1. This fusion gene involves two novel exons (in red box) upstream of ABCA5. (D) A selection of fusion long read alignments (bright blue) contributing structural information to isoform reconstruction is displayed in line with four reference annotations libraries: UCSC, RefSeq, GENCODE and Ensembl. The long read fragment alignments to UNK gene locus are shown at the left and the ones to ABCA5 gene locus at the right. The long read alignments detected two novel exons (in red box), while they are not reported by the reference annotation libraries. Note that ABCA5 is transcribed from the reverse strand. To show the fusion gene in the correct order, the browser figure of the ABCA5 gene locus is flipped. Please refer to Figure 7 for a description of figure elements.
Figure 8. IDP-fusion detected and annotated fusion genes with novel gene and novel exons involved. (A) IDP-fusion detected the fusion gene between AIB1 and an unannotated region in chromosome 1. Eight fusion isoforms were estimated at the modest abundance (RPKM > 1) and contain seven novel exons (blue) annotated by IDP-fusion. (B) A selection of fusion long read alignments (bright blue) contributing structural information to isoform reconstruction is displayed in line with four reference annotations libraries: UCSC, RefSeq, GENCODE and Ensembl. The long read fragment alignments to AIB1 gene locus are shown at the left and the ones to the novel gene locus at the right. The long read alignments annotated a novel gene with seven exons not reported by the reference annotation libraries. (C) IDP-fusion detected the fusion gene UNK–ABCA5 and two corresponding fusion isoforms with RPKM > 1. This fusion gene involves two novel exons (in red box) upstream of ABCA5. (D) A selection of fusion long read alignments (bright blue) contributing structural information to isoform reconstruction is displayed in line with four reference annotations libraries: UCSC, RefSeq, GENCODE and Ensembl. The long read fragment alignments to UNK gene locus are shown at the left and the ones to ABCA5 gene locus at the right. The long read alignments detected two novel exons (in red box), while they are not reported by the reference annotation libraries. Note that ABCA5 is transcribed from the reverse strand. To show the fusion gene in the correct order, the browser figure of the ABCA5 gene locus is flipped. Please refer to Figure 7 for a description of figure elements.CC BY · 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

1 retracted

Research works printed but not linked: 1

  1. Kim, 2015, Best practices for whole genome de novo sequencing with long-read smrt sequencing

Cited by

See all 129

Metadata

Authors
Weirather, J. L., Afshar, P. T., Clark, T. A., Tseng, E., Powers, L. S., Underwood, J. G., Zabner, J., Korlach, J., Wong, W. H., & Au, K. F.
Affiliations
  • University of Iowa (US)
  • Stanford University (US)
  • Pacific Biosciences (United States) (US)
  • University of Washington (US)
Corresponding author
Au, Kin Fai
Venue
Nucleic Acids Research
Published
2015-06-03
Volume
43
Issue
18
Pages
e116-e116
ISSN
  • 0305-1048 (ISSN-L)
  • 1362-4962
OpenAlex
W1944880632
Cited by
123
References
60
Citations by year
2026: 4 · 2025: 4 · 2024: 8 · 2023: 5 · 2022: 10 · 2021: 6 · 2020: 30 · 2019: 21 · 2018: 8 · 2017: 10 · 2016: 13 · 2015: 3
FWCI
4.92
Type
journal article
Language
en
Open access
Locations
  • Nucleic Acids Research (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
  • National Institutes of Health
  • National Human Genome Research Institute (R01HG007834)
  • National Institute of General Medical Sciences (R01GM109836)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Cancer Research › Cancer Genomics and Diagnostics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genomic variations and chromosomal abnormalities
Keywords
fusion genes, breast cancer, hybrid sequencing, long-read sequencing, short-read sequencing, MCF-7 cells
MeSH terms
Breast Neoplasms, Female, Humans, Sequence Alignment, Protein Isoforms, Gene Expression Profiling, Gene Fusion, High-Throughput Nucleotide Sequencing, MCF-7 Cells, Carcinogenesis

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.