CoChat

IW-Scoring: an Integrative Weighted Scoring framework for annotating and prioritizing genetic variations in the noncoding genome

Open access
Wang, Jun, Dayem Ullah, Abu Z, Chelala, Claude · 2018 · Nucleic Acids Research

Published version of a preprint

The DOI matches its Crossref record; 50 of the 50 references it cites were found in a registry or an index.

Abstract

The vast majority of germline and somatic variations occur in the noncoding part of the genome, only a small fraction of which are believed to be functional. From the tens of thousands of noncoding variations detectable in each genome, identifying and prioritizing driver candidates with putative functional significance is challenging. To address this, we implemented IW-Scoring, a new Integrative Weighted Scoring model to annotate and prioritise functionally relevant noncoding variations. We evaluate 11 scoring methods, and apply an unsupervised spectral approach for subsequent selective integration into two linear weighted functional scoring schemas for known and novel variations. IW-Scoring produces stable high-quality performance as the best predictors for three independent data sets. We demonstrate the robustness of IW-Scoring in identifying recurrent functional mutations in the TERT promoter, as well as disease SNPs in proximity to consensus motifs and with gene regulatory effects. Using follicular lymphoma as a paradigmatic cancer model, we apply IW-Scoring to locate 11 recurrently mutated noncoding regions in 14 follicular lymphoma genomes, and validate 9 of these regions in an extension cohort, including the promoter and enhancer regions of PAX5. Overall, IW-Scoring demonstrates greater versatility in identifying trait- and disease-associated noncoding variants. Scores from IW-Scoring as well as other methods are freely available from http://www.snp-nexus.org/IW-Scoring/.

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/gky057

Cites

50 of 50 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1. The IW-Scoring framework.
Figure 1. The IW-Scoring framework.CC BY · PubMed Central
Figure 2. Correlations between 11 functional scores and the distribution of integrative scores of 712 259 noncoding variants in the training set. (A) Correlation matrix of 11 functional scores based on variants in the training set. (B) Distribution of integrative scores based on 11 scoring systems for all variants in the training set. The mean and standard deviation were shown. Mean and median values were indicated by the blue and green lines, respectively.
Figure 2. Correlations between 11 functional scores and the distribution of integrative scores of 712 259 noncoding variants in the training set. (A) Correlation matrix of 11 functional scores based on variants in the training set. (B) Distribution of integrative scores based on 11 scoring systems for all variants in the training set. The mean and standard deviation were shown. Mean and median values were indicated by the blue and green lines, respectively.CC BY · PubMed Central
Figure 3. Comparison of the performances of different functional scores with IW-Scoring using the ClinVar pathogenic and benign noncoding variants. (A) ROC curves of IW-Scoring and 11 individual functional scores with AUC values noted, measuring the accuracy to differentiate pathogenic and benign variants. (B) AUC values for IW-Scoring and selected individual scores to differentiate pathogenic and benign variants. 95% CI bars were also shown. (C) Violin plots of integrative scores (without fitCons) for ClinVar pathogenic and benign variants. (D) Boxplot of P-values associated with integrative scores (without fitCons) for pathogenic and benign variants.
Figure 3. Comparison of the performances of different functional scores with IW-Scoring using the ClinVar pathogenic and benign noncoding variants. (A) ROC curves of IW-Scoring and 11 individual functional scores with AUC values noted, measuring the accuracy to differentiate pathogenic and benign variants. (B) AUC values for IW-Scoring and selected individual scores to differentiate pathogenic and benign variants. 95% CI bars were also shown. (C) Violin plots of integrative scores (without fitCons) for ClinVar pathogenic and benign variants. (D) Boxplot of P-values associated with integrative scores (without fitCons) for pathogenic and benign variants.CC BY · PubMed Central
Figure 4. Comparison of the performances of different functional scores using the GWAS and COSMIC noncoding data sets. (A) AUC values for IW-Scoring and individual scores in the ability to differentiate between GWAS and randomly selected 1000G noncoding SNPs. 95% CI bars were also shown. (B) AUC values for IW-Scoring and individual scores in the ability to differentiate COSMIC recurrent from non-recurrent noncoding mutations. (C) Violin plots of functional scores between recurrent and non-recurrent noncoding mutations for CADD, integrative scores (six scores) and ReMM.
Figure 4. Comparison of the performances of different functional scores using the GWAS and COSMIC noncoding data sets. (A) AUC values for IW-Scoring and individual scores in the ability to differentiate between GWAS and randomly selected 1000G noncoding SNPs. 95% CI bars were also shown. (B) AUC values for IW-Scoring and individual scores in the ability to differentiate COSMIC recurrent from non-recurrent noncoding mutations. (C) Violin plots of functional scores between recurrent and non-recurrent noncoding mutations for CADD, integrative scores (six scores) and ReMM.CC BY · PubMed Central
Figure 5. Application of IW-Scoring in scoring variants identified in GWAS, eQTL and cancer studies. (A) Disease SNPs near consensus motifs in SMAD3 intronic region. Integrative scores (IW-score K10) were shown for all nearby known SNPs along with the disease-associated candidates. The score of rs17293632 was compared to the mean value and standard deviation of all nearby known variants, and P-value was derived and shown for each method. The mean difference in scores between rs17293632 and all other nearby SNPs after rescaling (mean of zero and SD of one, for the convenience of cross-comparison) for each method was also shown at the right panel. (B) Disease SNPs with gene regulatory effects nearly 3′ UTR of IKZF3. Integrative scores (IW-score K10) were shown for 1,000 randomly selected known SNPs nearby, and two GWAS and eQTL candidates, circled with red and blue circles, respectively. Functional scores of rs12946510 and rs907091 for all methods were also extracted and compared, rescaled to fit into the rescaled distribution of the training set. The plots were shown at the right panel. (C) TERT promoter mutations and known SNPs in the region, with their integrative scores calculated. Black and grey dots denote somatic mutations and SNPs, respectively. Functional scores between TERT promoter (defined by the DNase I mark of the GM12878 cell line) and non-promoter regions were compared and P-values were derived at the right panel.
Figure 5. Application of IW-Scoring in scoring variants identified in GWAS, eQTL and cancer studies. (A) Disease SNPs near consensus motifs in SMAD3 intronic region. Integrative scores (IW-score K10) were shown for all nearby known SNPs along with the disease-associated candidates. The score of rs17293632 was compared to the mean value and standard deviation of all nearby known variants, and P-value was derived and shown for each method. The mean difference in scores between rs17293632 and all other nearby SNPs after rescaling (mean of zero and SD of one, for the convenience of cross-comparison) for each method was also shown at the right panel. (B) Disease SNPs with gene regulatory effects nearly 3′ UTR of IKZF3. Integrative scores (IW-score K10) were shown for 1,000 randomly selected known SNPs nearby, and two GWAS and eQTL candidates, circled with red and blue circles, respectively. Functional scores of rs12946510 and rs907091 for all methods were also extracted and compared, rescaled to fit into the rescaled distribution of the training set. The plots were shown at the right panel. (C) TERT promoter mutations and known SNPs in the region, with their integrative scores calculated. Black and grey dots denote somatic mutations and SNPs, respectively. Functional scores between TERT promoter (defined by the DNase I mark of the GM12878 cell line) and non-promoter regions were compared and P-values were derived at the right panel.CC BY · PubMed Central
Figure 6. Integrative scores IW-score (N8) of somatic mutations in follicular lymphoma. (A) Integrative scores for mutations within Ensembl Regulatory Build annotated regions. (B) Integrative scores for mutations within top mutated TF binding sites. The mean value was shown by the red dotted line. (C) Recurrent mutations in PAX5 locus. Integrative scores were shown for all noncoding mutations in whole chromosome 9 first. PAX5 locus was further zoomed in with three mutations clusters further identified, PAX5 promoter, putative enhance regions 1 and 2. Mutations from ICGC FL samples were shown by solid green dots, while mutations from Okosun et al. were shown by solid light blue dots. Signals of epigenetic histone marks were for GM12878 cell line. (D) Integrative scores for mutations across three PAX5 associated regions in boxplots. Red dotted line indicated the level of P = 0.05.
Figure 6. Integrative scores IW-score (N8) of somatic mutations in follicular lymphoma. (A) Integrative scores for mutations within Ensembl Regulatory Build annotated regions. (B) Integrative scores for mutations within top mutated TF binding sites. The mean value was shown by the red dotted line. (C) Recurrent mutations in PAX5 locus. Integrative scores were shown for all noncoding mutations in whole chromosome 9 first. PAX5 locus was further zoomed in with three mutations clusters further identified, PAX5 promoter, putative enhance regions 1 and 2. Mutations from ICGC FL samples were shown by solid green dots, while mutations from Okosun et al. were shown by solid light blue dots. Signals of epigenetic histone marks were for GM12878 cell line. (D) Integrative scores for mutations across three PAX5 associated regions in boxplots. Red dotted line indicated the level of P = 0.05.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

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 36

Metadata

Authors
Wang, J., Dayem Ullah, A. Z., & Chelala, C.
Affiliations
  • Queen Mary University of London (GB)
  • Centre for Molecular Oncology, Barts Cancer Institute, Queen Mary University of London, Charterhouse Square, London EC1M 6BQ, UK
Corresponding author
Wang, Jun
Venue
Nucleic Acids Research
Published
2018-01-30
Volume
46
Issue
8
Pages
e47-e47
ISSN
  • 0305-1048 (ISSN-L)
  • 1362-4962
OpenAlex
W2951354947
Cited by
35
References
50
Citations by year
2026: 2 · 2025: 4 · 2024: 1 · 2023: 4 · 2022: 6 · 2021: 6 · 2020: 6 · 2019: 5 · 2018: 1
FWCI
2.75
Type
journal article
Language
en
Open access
Locations
  • Nucleic Acids Research (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc-nd)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Queen Mary Research Online (Queen Mary University of London) (submittedVersion, cc-by)
Funders
  • Cancer Research UK (C16420/A18066)
  • Barts and The London Charity (MGU0344)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genomics and Rare Diseases
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Associations and Epidemiology
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Chromatin Dynamics
Keywords
non-coding variants, variant prioritization, variant annotation, weighted scoring model, spectral clustering, TERT promoter mutation, follicular lymphoma, recurrent mutations, disease-associated SNPs, regulatory motifs, somatic mutations
MeSH terms
Whole Genome Sequencing, Humans, Lymphoma, Follicular, Models, Genetic, Mutation, Neoplasms, Promoter Regions, Genetic, Regulatory Sequences, Nucleic Acid, Genetic Variation, Genome, Human, Telomerase, Computational Biology, Polymorphism, Single Nucleotide, DNA, Intergenic, Databases, Nucleic Acid, Genome-Wide Association Study

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.