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

The DOI matches its Crossref record; 68 of the 70 references it cites were found in a registry or an index.

Paralog buffering contributes to the variable essentiality of genes in cancer cell lines

Open access
De Kegel, Barbara, Ryan, Colm J. · 2019 · PLOS Genetics

Published version of a preprint

Abstract

What makes a gene essential for cellular survival? In model organisms, such as budding yeast, systematic gene deletion studies have revealed that paralog genes are less likely to be essential than singleton genes and that this can partially be attributed to the ability of paralogs to buffer each other's loss. However, the essentiality of a gene is not a fixed property and can vary significantly across different genetic backgrounds. It is unclear to what extent paralogs contribute to this variation, as most studies have analyzed genes identified as essential in a single genetic background. Here, using gene essentiality profiles of 558 genetically heterogeneous tumor cell lines, we analyze the contribution of paralogy to variable essentiality. We find that, compared to singleton genes, paralogs are less frequently essential and that this is more evident when considering genes with multiple paralogs or with highly sequence-similar paralogs. In addition, we find that paralogs derived from whole genome duplication exhibit more variable essentiality than those derived from small-scale duplications. We provide evidence that in 13-17% of cases the variable essentiality of paralogs can be attributed to buffering relationships between paralog pairs, as evidenced by synthetic lethality. Paralog pairs derived from whole genome duplication and pairs that function in protein complexes are significantly more likely to display such synthetic lethal relationships. Overall we find that many of the observations made using a single strain of budding yeast can be extended to understand patterns of essentiality in genetically heterogeneous cancer cell lines.

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 · 10.1371/journal.pgen.1008466

Cites

68 of 68 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig 1 Obtaining essentiality data for 16,540 protein-coding genes in 558 cancer cell lines. (A) Workflow showing how gene scores were pre-processed and subsequently binarized. (B) Histogram showing the distribution of all gene scores across all cell lines. Curves show the 3-components of the Gaussian mixture model fit to this distribution. Colors indicate the component model each region of the histogram was assigned to (differing shades of blue for different components). The threshold (-0.47) used to convert gene scores from continuous values to binary essential and non-essential calls is shown using the dotted line.
Fig 1 Obtaining essentiality data for 16,540 protein-coding genes in 558 cancer cell lines. (A) Workflow showing how gene scores were pre-processed and subsequently binarized. (B) Histogram showing the distribution of all gene scores across all cell lines. Curves show the 3-components of the Gaussian mixture model fit to this distribution. Colors indicate the component model each region of the histogram was assigned to (differing shades of blue for different components). The threshold (-0.47) used to convert gene scores from continuous values to binary essential and non-essential calls is shown using the dotted line.CC BY · PubMed Central
Fig 2 Genes that are never or rarely essential are enriched in paralogs. (A) Distribution of the percentage of cell lines in which a gene is essential. The colors indicate the three broad categories of essentiality: genes that are essential in 0% of cell lines (blue), genes that are essential in more than 0 but at most 90% of cell lines (gray), and genes that are essential in 90% or more of the cell lines (orange). (B) Stacked bar graph showing the proportion of genes in each of the three essentiality categories that are paralogs (yellow) vs. those that are singletons (cyan). (C) Stacked bar graph showing, for all genes binned according to the percentage of cell lines in which they are essential, the proportion of genes in each bin that are paralogs (yellow), or singletons (cyan). For reference, on the right is a bar showing the proportion of genes in the full dataset that are paralogs or singletons.
Fig 2 Genes that are never or rarely essential are enriched in paralogs. (A) Distribution of the percentage of cell lines in which a gene is essential. The colors indicate the three broad categories of essentiality: genes that are essential in 0% of cell lines (blue), genes that are essential in more than 0 but at most 90% of cell lines (gray), and genes that are essential in 90% or more of the cell lines (orange). (B) Stacked bar graph showing the proportion of genes in each of the three essentiality categories that are paralogs (yellow) vs. those that are singletons (cyan). (C) Stacked bar graph showing, for all genes binned according to the percentage of cell lines in which they are essential, the proportion of genes in each bin that are paralogs (yellow), or singletons (cyan). For reference, on the right is a bar showing the proportion of genes in the full dataset that are paralogs or singletons.CC BY · PubMed Central
Fig 3 Genes with more paralogs and genes which share greater sequence identity with their closest paralog are less frequently essential. (A) Bar chart showing counts of the number of paralogs, from zero to four or more, for all protein-coding genes. (B) Stacked bar graph showing, for genes binned according to the number of paralogs they have, the proportion of genes in each bin that are never, sometimes and broadly essential. (C) The kernel density estimates of the percent of a gene’s protein sequence that is identical in its closest paralog, for paralogs in each essentiality category. Drawn using Seaborn’s kdeplot function with the default parameters.
Fig 3 Genes with more paralogs and genes which share greater sequence identity with their closest paralog are less frequently essential. (A) Bar chart showing counts of the number of paralogs, from zero to four or more, for all protein-coding genes. (B) Stacked bar graph showing, for genes binned according to the number of paralogs they have, the proportion of genes in each bin that are never, sometimes and broadly essential. (C) The kernel density estimates of the percent of a gene’s protein sequence that is identical in its closest paralog, for paralogs in each essentiality category. Drawn using Seaborn’s kdeplot function with the default parameters.CC BY · PubMed Central
Fig 4 WGDs are more variably essential than SSDs. (A) Stacked bar graph showing the proportions of WGDs, SSDs and singletons that are never, sometimes and broadly essential. (B) Bar graph showing the proportion of WGDs, SSDs and singletons that are broadly essential (zoomed in version of the orange portion in A).
Fig 4 WGDs are more variably essential than SSDs. (A) Stacked bar graph showing the proportions of WGDs, SSDs and singletons that are never, sometimes and broadly essential. (B) Bar graph showing the proportion of WGDs, SSDs and singletons that are broadly essential (zoomed in version of the orange portion in A).CC BY · PubMed Central
Fig 5 Variation in paralog expression explains why some paralogs are variably essential. (A) Top: Workflow showing selection of paralog pairs (A1, A2). Bottom: Boxplot illustrating the hypothesis that was tested for each of these pairs, namely that the expression of A2 would be higher (on average) in the cell lines where A1 is not essential compared to the cell lines where A1 is essential. (B) Scatterplot of all tested pairs showing the mean expression of A2 in the cell lines where A1 is not essential (x-axis) or essential (y-axis). Colored dots indicate pairs for which the expression of A2 is significantly lower in the cell lines where A1 is essential, i.e. putative synthetic lethal pairs. Grey dots represent pairs that were tested but for which no significant difference in expression was found between cell lines where A1 is and is not essential. Color corresponds to duplication type—pink for WGDs and green for SSDs. Selected synthetic lethalities involving protein complex subunits are labelled. (C) Boxplots showing gene pairs (e.g. A1 = SMARCA4, A2 = SMARCA2) for which A2 expression is on average higher in the cell lines where A1 is not essential compared to cell lines where A1 is essential.
Fig 5 Variation in paralog expression explains why some paralogs are variably essential. (A) Top: Workflow showing selection of paralog pairs (A1, A2). Bottom: Boxplot illustrating the hypothesis that was tested for each of these pairs, namely that the expression of A2 would be higher (on average) in the cell lines where A1 is not essential compared to the cell lines where A1 is essential. (B) Scatterplot of all tested pairs showing the mean expression of A2 in the cell lines where A1 is not essential (x-axis) or essential (y-axis). Colored dots indicate pairs for which the expression of A2 is significantly lower in the cell lines where A1 is essential, i.e. putative synthetic lethal pairs. Grey dots represent pairs that were tested but for which no significant difference in expression was found between cell lines where A1 is and is not essential. Color corresponds to duplication type—pink for WGDs and green for SSDs. Selected synthetic lethalities involving protein complex subunits are labelled. (C) Boxplots showing gene pairs (e.g. A1 = SMARCA4, A2 = SMARCA2) for which A2 expression is on average higher in the cell lines where A1 is not essential compared to cell lines where A1 is essential.CC BY · PubMed Central
Fig 6 Duplication mode, protein complex membership and sequence similarity are predictive of synthetic lethality. (A) Bar chart showing the percentage of tested WGD and SSD pairs that are synthetic lethal. (B) Bar chart showing the percentage of synthetic lethal pairs observed among the tested pairs where at least one gene is a protein complex member versus the tested pairs where neither gene is a protein complex member. (C) Same as (B) but restricted to tested pairs that are WGDs. (D) Same as (A) but restricted to tested pairs where at least one gene is a protein complex member. (E) Boxplots showing the percent of the protein sequence that is identical in the members of a given paralog pair, for pairs that are synthetic lethal versus those that are not. The pairs used in the analysis consist of all sometimes essential paralogs and their most sequence-similar paralog, and three subsets thereof. From left to right the subsets consist of: pairs that are WGDs, pairs that are SSDs, and pairs where either gene is a protein complex member.
Fig 6 Duplication mode, protein complex membership and sequence similarity are predictive of synthetic lethality. (A) Bar chart showing the percentage of tested WGD and SSD pairs that are synthetic lethal. (B) Bar chart showing the percentage of synthetic lethal pairs observed among the tested pairs where at least one gene is a protein complex member versus the tested pairs where neither gene is a protein complex member. (C) Same as (B) but restricted to tested pairs that are WGDs. (D) Same as (A) but restricted to tested pairs where at least one gene is a protein complex member. (E) Boxplots showing the percent of the protein sequence that is identical in the members of a given paralog pair, for pairs that are synthetic lethal versus those that are not. The pairs used in the analysis consist of all sometimes essential paralogs and their most sequence-similar paralog, and three subsets thereof. From left to right the subsets consist of: pairs that are WGDs, pairs that are SSDs, and pairs where either gene is a protein complex member.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

Show all 70 references

Research works printed but not linked: 3

  1. E Jones, 2001, SciPy: Open source scientific tools for Python
  2. M Morgan, 2016, Rsamtools: Binary alignment (BAM), FASTA, variant call (BCF), and tabix file import, R package version, 1, 677
  3. H Pagès, 2017, BSgenome: Software infrastructure for efficient representation of full genomes and their SNPs, R package version, 1, 10

Cited by

See all 94

Metadata

Authors
De Kegel, B., & Ryan, C. J.
Affiliations
University College Dublin (IE)
Corresponding author
Ryan, Colm J.
Venue
PLOS Genetics
Published
2019-10-25
Volume
15
Issue
10
Pages
e1008466
ISSN
  • 1553-7390 (ISSN-L)
  • 1553-7404
OpenAlex
W2981480342
Cited by
92
References
70
Citations by year
2026: 8 · 2025: 14 · 2024: 11 · 2023: 14 · 2022: 18 · 2021: 14 · 2020: 13
FWCI
3.56
Type
journal article
Language
en
Open access
Locations
  • PLoS Genetics (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc-nd)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Figshare (submittedVersion, cc-by)
Funders
Irish Research Council (Laureate Awards 2017/2018)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Cancer Research › Cancer Genomics and Diagnostics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Evolution and Genetic Dynamics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Bioinformatics and Genomic Networks
Keywords
gene essentiality, paralogs, synthetic lethality, whole-genome duplication, cancer cell lines, genetic heterogeneity, protein complexes, sequence similarity
MeSH terms
Synthetic Lethal Mutations, Saccharomycetales, Humans, Models, Genetic, Neoplasms, Gene Deletion, Evolution, Molecular, Genes, Essential, Gene Duplication, Cell Line, Tumor

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

How to cite

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

This page
CoChat. (2026). Citation check: Paralog buffering contributes to the variable essentiality of genes in cancer cell lines. https://app.cochat.ai/papers/10.1371/journal.pgen.1008466
The paper
De Kegel, B., & Ryan, C. J. (2019). Paralog buffering contributes to the variable essentiality of genes in cancer cell lines. PLOS Genetics. https://doi.org/10.1371/journal.pgen.1008466

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.