CoChat
Sign Up
Cited by 640Genomics and Rare DiseasesRNA Research and SplicingGenomics and Phylogenetic Studies

CADD-Splice—improving genome-wide variant effect prediction using deep learning-derived splice scores

Open access
Rentzsch, Philipp, Schubach, Max, Shendure, Jay, Kircher, Martin·2021·Genome Medicine
DOI 10.1186/s13073-021-00835-9PMC PMC7901104
Open this paper in CoChat — free accountOpen PDF

Abstract

Background: Splicing of genomic exons into mRNAs is a critical prerequisite for the accurate synthesis of human proteins. Genetic variants impacting splicing underlie a substantial proportion of genetic disease, but are challenging to identify beyond those occurring at donor and acceptor dinucleotides. To address this, various methods aim to predict variant effects on splicing. Recently, deep neural networks (DNNs) have been shown to achieve better results in predicting splice variants than other strategies. Methods: It has been unclear how best to integrate such process-specific scores into genome-wide variant effect predictors. Here, we use a recently published experimental data set to compare several machine learning methods that score variant effects on splicing. We integrate the best of those approaches into general variant effect prediction models and observe the effect on classification of known pathogenic variants. Results: We integrate two specialized splicing scores into CADD (Combined Annotation Dependent Depletion; cadd.gs.washington.edu ), a widely used tool for genome-wide variant effect prediction that we previously developed to weight and integrate diverse collections of genomic annotations. With this new model, CADD-Splice, we show that inclusion of splicing DNN effect scores substantially improves predictions across multiple variant categories, without compromising overall performance. Conclusions: While splice effect scores show superior performance on splice variants, specialized predictors cannot compete with other variant scores in general variant interpretation, as the latter account for nonsense and missense effects that do not alter splicing. Although only shown here for splice scores, we believe that the applied approach will generalize to other specific molecular processes, providing a path for the further improvement of genome-wide variant effect prediction.

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 · 10.1186/s13073-021-00835-9

Cites

6 of 6 references verified0 citation errors · 0 not checkable

Statistics

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?

  • All variants were prioritized considering allele frequency in gnomAD v4.0.0 below 1%, impact on protein function via different in silico programs, and involvement of candidate genes in neuronal processes.

    “All variants were prioritized considering allele frequency in gnomAD v4.0.0 (Karczewski et al. 2020) below 1%, impact on protein function via different in silico programs (CADD (Rentzsch et al. 2021), REVEL (Ioannidis et al. 2016), MutPred2 (Pejaver et al. 2020) and Metadome (Wiel et al. 2019)) and involvement of candidate genes in neuronal processes.”

    From Heterozygous loss-of-function variants in DOCK4 cause neurodevelopmental delay and microcephaly (2024) · cites this paper as Rentzsch et al. 2021

Self-check

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

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

Cites

Flagged first, then by citations

Research works printed but not linked: 52

  1. [1]Sibley CR Blazquez L Ule J Lessons from non-canonical splicing Nat Rev Genet 2016 17 407 421 10.1038/nrg.2016.46 27240813 PMC5154377 · DOI
  2. [2]Baralle FE Giudice J Alternative splicing as a regulator of development and tissue identity Nat Rev Mol Cell Biol 2017 18 437 451 10.1038/nrm.2017.27 28488700 PMC6839889 · DOI · Paper page
  3. [3]Wang ET Sandberg R Luo S Alternative isoform regulation in human tissue transcriptomes Nature 2008 456 470 476 10.1038/nature07509 18978772 PMC2593745 · DOI · Paper page
  4. [4]Pan Q Shai O Lee LJ Deep surveying of alternative splicing complexity in the human transcriptome by high-throughput sequencing Nat Genet 2008 40 1413 1415 10.1038/ng.259 18978789 · DOI · Paper page
  5. [5]Katz Y Wang ET Airoldi EM Burge CB Analysis and design of RNA sequencing experiments for identifying isoform regulation Nat Methods 2010 7 1009 1015 10.1038/nmeth.1528 21057496 PMC3037023 · DOI · Paper page
  6. [6]Cummings BB Karczewski KJ Kosmicki JA Transcript expression-aware annotation improves rare variant interpretation Nature 2020 581 452 458 10.1038/s41586-020-2329-2 32461655 PMC7334198 · DOI
  7. [7]Melé M Ferreira PG Reverter F The human transcriptome across tissues and individuals Science 2015 348 660 665 10.1126/science.aaa0355 25954002 PMC4547472 · DOI · Paper page
  8. [8]Li YI van de Geijn B Raj A RNA splicing is a primary link between genetic variation and disease Science 2016 352 600 604 10.1126/science.aad9417 27126046 PMC5182069 · DOI · Paper page
  9. [9]Scotti MM Swanson MS RNA mis-splicing in disease Nat Rev Genet 2016 17 19 32 10.1038/nrg.2015.3 26593421 PMC5993438 · DOI
  10. [10]Li X Kim Y Tsang EK The impact of rare variation on gene expression across tissues Nature 2017 550 239 243 10.1038/nature24267 29022581 PMC5877409 · DOI
  11. [12]Buske OJ Manickaraj A Mital S Identification of deleterious synonymous variants in human genomes Bioinforma Oxf Engl 2013 29 1843 1850 10.1093/bioinformatics/btt308 23736532 · DOI
  12. [13]Vaser R Adusumalli S Leng SN SIFT missense predictions for genomes Nat Protoc 2016 11 1 9 10.1038/nprot.2015.123 26633127 · DOI · Paper page
  13. [14]Adzhubei I, Jordan DM, Sunyaev SR (2013) Predicting functional effect of human missense mutations using PolyPhen-2. Curr Protoc Hum Genet chapter 7:Unit7.20. doi: 10.1002/0471142905.hg0720s76. PMC4480630 23315928 · DOI · Paper page
  14. [15]Liu X Wu C Li C Boerwinkle E dbNSFP v3.0: a one-stop database of functional predictions and annotations for human nonsynonymous and splice-site SNVs Hum Mutat 2016 37 235 241 10.1002/humu.22932 26555599 PMC4752381 · DOI
  15. [16]Hu J Ng PC Predicting the effects of frameshifting indels Genome Biol 2012 13 R9 10.1186/gb-2012-13-2-r9 22322200 PMC3334572 · DOI
  16. [17]Pagel KA Pejaver V Lin GN When loss-of-function is loss of function: assessing mutational signatures and impact of loss-of-function genetic variants Bioinformatics 2017 33 i389 i398 10.1093/bioinformatics/btx272 28882004 PMC5870554 · DOI
  17. [18]Zhou J Troyanskaya OG Predicting effects of noncoding variants with deep learning-based sequence model Nat Methods 2015 12 931 934 10.1038/nmeth.3547 26301843 PMC4768299 · DOI · Paper page
  18. [19]di Iulio J Bartha I Wong EHM The human noncoding genome defined by genetic diversity Nat Genet 2018 50 333 337 10.1038/s41588-018-0062-7 29483654 · DOI · Paper page
  19. [20]Rentzsch P Witten D Cooper GM CADD: predicting the deleteriousness of variants throughout the human genome Nucleic Acids Res 2019 47 D886 D894 10.1093/nar/gky1016 30371827 PMC6323892 · DOI · Paper page
  20. [21]Ionita-Laza I McCallum K Xu B Buxbaum JD A spectral approach integrating functional genomic annotations for coding and noncoding variants Nat Genet 2016 48 214 220 10.1038/ng.3477 26727659 PMC4731313 · DOI
  21. [22]Shihab HA Rogers MF Gough J An integrative approach to predicting the functional effects of non-coding and coding sequence variation Bioinforma Oxf Engl 2015 31 1536 1543 10.1093/bioinformatics/btv009 PMC4426838 25583119 · DOI
  22. [23]Kircher M Witten DM Jain P A general framework for estimating the relative pathogenicity of human genetic variants Nat Genet 2014 46 310 315 10.1038/ng.2892 24487276 PMC3992975 · DOI · Paper page
  23. [25]Mather CA Mooney SD Salipante SJ CADD score has limited clinical validity for the identification of pathogenic variants in noncoding regions in a hereditary cancer panel Genet Med 2016 18 1269 1275 10.1038/gim.2016.44 27148939 PMC5097698 · DOI
  24. [26]Cheung R Insigne KD Yao D A multiplexed assay for exon recognition reveals that an unappreciated fraction of rare genetic variants cause large-effect splicing disruptions Mol Cell 2019 73 183 194 10.1016/j.molcel.2018.10.037 30503770 PMC6599603 · DOI · Paper page
  25. [28]Pollard KS Hubisz MJ Rosenbloom KR Siepel A Detection of nonneutral substitution rates on mammalian phylogenies Genome Res 2010 20 110 121 10.1101/gr.097857.109 19858363 PMC2798823 · DOI · Paper page
  26. [29]Jian X Boerwinkle E Liu X In silico tools for splicing defect prediction - a survey from the viewpoint of end-users Genet Med Off J Am Coll Med Genet 2014 16 497 10.1038/gim.2013.176 PMC4029872 24263461 · DOI · Paper page
  27. [30]Anna A Monika G Splicing mutations in human genetic disorders: examples, detection, and confirmation J Appl Genet 2018 59 253 268 10.1007/s13353-018-0444-7 29680930 PMC6060985 · DOI · Paper page
  28. [31]Mort M Sterne-Weiler T Li B MutPred splice: machine learning-based prediction of exonic variants that disrupt splicing Genome Biol 2014 15 R19 10.1186/gb-2014-15-1-r19 24451234 PMC4054890 · DOI
  29. [32]Jagadeesh KA Paggi JM Ye JS S-CAP extends pathogenicity prediction to genetic variants that affect RNA splicing Nat Genet 2019 51 755 763 10.1038/s41588-019-0348-4 30804562 · DOI
  30. [33]Jian X Boerwinkle E Liu X In silico prediction of splice-altering single nucleotide variants in the human genome Nucleic Acids Res 2014 42 13534 13544 10.1093/nar/gku1206 25416802 PMC4267638 · DOI
  31. [34]Soemedi R Cygan KJ Rhine CL Pathogenic variants that alter protein code often disrupt splicing Nat Genet 2017 49 848 855 10.1038/ng.3837 28416821 PMC6679692 · DOI
  32. [35]Yeo G Burge CB Maximum entropy modeling of short sequence motifs with applications to RNA splicing signals J Comput Biol J Comput Mol Cell Biol 2004 11 377 394 10.1089/1066527041410418 15285897 · DOI
  33. [36]Ke S Shang S Kalachikov SM Quantitative evaluation of all hexamers as exonic splicing elements Genome Res 2011 21 1360 1374 10.1101/gr.119628.110 21659425 PMC3149502 · DOI
  34. [37]Rosenberg AB Patwardhan RP Shendure J Seelig G Learning the sequence determinants of alternative splicing from millions of random sequences Cell 2015 163 698 711 10.1016/j.cell.2015.09.054 26496609 · DOI
  35. [38]Reese MG Eeckman FH Kulp D Haussler D Improved splice site detection in genie J Comput Biol J Comput Mol Cell Biol 1997 4 311 323 10.1089/cmb.1997.4.311 9278062 · DOI
  36. [39]Xiong HY Alipanahi B Lee LJ RNA splicing. The human splicing code reveals new insights into the genetic determinants of disease Science 2015 347 1254806 10.1126/science.1254806 25525159 PMC4362528 · DOI
  37. [40]Cheng J Nguyen TYD Cygan KJ MMSplice: modular modeling improves the predictions of genetic variant effects on splicing Genome Biol 2019 20 48 10.1186/s13059-019-1653-z 30823901 PMC6396468 · DOI · Paper page
  38. [41]Jaganathan K Panagiotopoulou SK McRae JF Predicting splicing from primary sequence with deep learning Cell 2019 176 414 416 10.1016/j.cell.2018.12.015 30661751 · DOI · Paper page
  39. [42]Lek M Karczewski KJ Minikel EV Analysis of protein-coding genetic variation in 60,706 humans Nature 2016 536 285 291 10.1038/nature19057 27535533 PMC5018207 · DOI · Paper page
  40. [43]Avsec Ž Kreuzhuber R Israeli J The Kipoi repository accelerates community exchange and reuse of predictive models for genomics Nat Biotechnol 2019 37 592 600 10.1038/s41587-019-0140-0 31138913 PMC6777348 · DOI
  41. [44]Aken BL, Ayling S, Barrell D, et al. The Ensembl gene annotation system. Database. 2016, 2016:baw093. 10.1093/database/baw093. PMC4919035 27337980 · DOI
  42. [46]Karczewski KJ Francioli LC Tiao G The mutational constraint spectrum quantified from variation in 141,456 humans Nature 2020 581 434 443 10.1038/s41586-020-2308-7 32461654 PMC7334197 · DOI · Paper page
  43. [47]Zhao H Sun Z Wang J CrossMap: a versatile tool for coordinate conversion between genome assemblies Bioinformatics 2014 30 1006 1007 10.1093/bioinformatics/btt730 24351709 PMC3967108 · DOI
  44. [48]McLaren W Gil L Hunt SE The Ensembl variant effect predictor Genome Biol 2016 17 122 10.1186/s13059-016-0974-4 27268795 PMC4893825 · DOI · Paper page
  45. [52]Shen S Park JW Huang J MATS: a Bayesian framework for flexible detection of differential alternative splicing from RNA-Seq data Nucleic Acids Res 2012 40 e61 10.1093/nar/gkr1291 22266656 PMC3333886 · DOI
  46. [53]Park E Pan Z Zhang Z The expanding landscape of alternative splicing variation in human populations Am J Hum Genet 2018 102 11 26 10.1016/j.ajhg.2017.11.002 29304370 PMC5777382 · DOI
  47. [54]Ling JP Wilks C Charles R ASCOT identifies key regulators of neuronal subtype-specific splicing Nat Commun 2020 11 137 10.1038/s41467-019-14020-5 31919425 PMC6952364 · DOI · Paper page
  48. [55]Baeza-Centurion P Miñana B Schmiedel JM Combinatorial genetics reveals a scaling law for the effects of mutations on splicing Cell 2019 176 549 563.e23 10.1016/j.cell.2018.12.010 30661752 · DOI
  49. [56]Anderson D Baynam G Blackwell JM Lassmann T Personalised analytics for rare disease diagnostics Nat Commun 2019 10 1 8 10.1038/s41467-019-13345-5 31754101 PMC6872807 · DOI · Paper page
  50. [57]Mohammadi P Castel SE Cummings BB Genetic regulatory variation in populations informs transcriptome analysis in rare disease Science 2019 366 351 356 10.1126/science.aay0256 31601707 PMC6814274 · DOI
  51. [58]Havrilla JM Pedersen BS Layer RM Quinlan AR A map of constrained coding regions in the human genome Nat Genet 2019 51 88 10.1038/s41588-018-0294-6 30531870 PMC6589356 · DOI
  52. [59]Abramovs N Brass A Tassabehji M GeVIR is a continuous gene-level metric that uses variant distribution patterns to prioritize disease candidate genes Nat Genet 2020 52 35 39 10.1038/s41588-019-0560-2 31873297 · DOI

Cited by

Sorted by citations

Authors
Rentzsch, P., Schubach, M., Shendure, J., & Kircher, M.
Affiliations
  • Berlin Institute of Health at Charité - Universitätsmedizin Berlin (DE)
  • University of Washington (US)
Corresponding author
Kircher, Martin
ORCID
Venue
Genome Medicine
Publisher
Springer Science and Business Media LLC
Published
2021-02-22
Volume
13
Issue
1
Pages
31-31
ISSN
1756-994X (ISSN-L)
DOI
10.1186/s13073-021-00835-9
PMC
PMC7901104
OpenAlex
W3129351841
Cited by
640
References
6
Citations by year
2026: 66 · 2025: 103 · 2024: 136 · 2023: 170 · 2022: 134 · 2021: 30 · 2020: 1
FWCI
67.99
Type
journal article
Language
en
Open access
Licence
Locations
  • Genome Medicine (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Refubium (Universitätsbibliothek der Freien Universität Berlin) (submittedVersion, cc-by)
  • Universitätsbibliothek der FU Berlin Hochschulschriftenstelle u. Dokumentenserver (cc-by)
Funders
  • National Cancer Institute (1R01CA197139)
  • Berlin Institute of Health
  • Charité
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genomics and Rare Diseases
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › RNA Research and Splicing
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
Keywords
variant effect prediction, splicing, CADD, pathogenic variant classification, splice variants, machine learning integration
MeSH terms
Deep Learning, Base Sequence, Exons, Humans, Introns, RNA Splicing, Genetic Variation, 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