A deep learning approach to automate refinement of somatic variant calling from cancer sequencing data

Cancer genomic analysis requires accurate identification of somatic variants in sequencing data. Manual review to refine somatic variant calls is required as a final step after automated processing. However, manual variant refinement is time-consuming, costly, poorly standardized, and non-reproducib...

Full description

Saved in:
Bibliographic Details
Published inNature genetics Vol. 50; no. 12; pp. 1735 - 1743
Main Authors Ainscough, Benjamin J., Barnell, Erica K., Ronning, Peter, Campbell, Katie M., Wagner, Alex H., Fehniger, Todd A., Dunn, Gavin P., Uppaluri, Ravindra, Govindan, Ramaswamy, Rohan, Thomas E., Griffith, Malachi, Mardis, Elaine R., Swamidass, S. Joshua, Griffith, Obi L.
Format Journal Article
LanguageEnglish
Published New York Nature Publishing Group US 01.12.2018
Nature Publishing Group
Subjects
45
631/114/1314
631/114/2785
631/114/794
631/208/514/2184
692/699/67
Agriculture
Algorithms
Animal Genetics and Genomics
Artificial intelligence
Automation
Biomedical and Life Sciences
Biomedicine
Breast cancer
Cancer
Cancer genetics
Cancer Research
Computer Simulation
Data processing
Datasets
Deep Learning
DNA Mutational Analysis - instrumentation
DNA Mutational Analysis - methods
Electronic Data Processing - methods
Gene Function
Genetic variation
Genomes
Genomic analysis
Genomics
Handbooks
High-Throughput Nucleotide Sequencing - instrumentation
Human Genetics
Humans
Learning algorithms
Leukemia
Machine learning
Mutation
Neoplasms - genetics
Polymorphism, Single Nucleotide
Quality standards
Reproducibility of Results
Sequence Analysis, DNA - instrumentation
Sequence Analysis, DNA - methods
Software
technical-report
Technology application
Tumors
Online AccessGet full text

Cover

More Information
Summary:Cancer genomic analysis requires accurate identification of somatic variants in sequencing data. Manual review to refine somatic variant calls is required as a final step after automated processing. However, manual variant refinement is time-consuming, costly, poorly standardized, and non-reproducible. Here, we systematized and standardized somatic variant refinement using a machine learning approach. The final model incorporates 41,000 variants from 440 sequencing cases. This model accurately recapitulated manual refinement labels for three independent testing sets (13,579 variants) and accurately predicted somatic variants confirmed by orthogonal validation sequencing data (212,158 variants). The model improves on manual somatic refinement by reducing bias on calls otherwise subject to high inter-reviewer variability. A machine learning approach for refinement of somatic variant calls automates this process and reduces bias stemming from inter-reviewer variability.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ObjectType-Article-2
ObjectType-Undefined-1
ObjectType-Feature-3
content type line 23
Author contributions B.J.A. designed the study, assembled and cleaned training data, performed feature engineering, designed model architecture, tuned hyperparameters, performed model training and analysis, performed manual review, assembled validation data, wrote code, created figures, and wrote the manuscript. E.K.B. designed the study, performed manual review, performed model training and analysis, performed clinical data analysis, assembled validation data, wrote code, created figures, and wrote the manuscript. P.R. and K.M.C. wrote code, performed manual review, and edited the manuscript. A.H.W. wrote code. T.E.R., R.G., R.U., G.P.D, and T.A.F. shared genomic data that was used in training the model and revised the paper. M.G., E.R.M., S.J.S., and O.L.G. designed the study, supervised the project and revised the paper.
ISSN:1061-4036
1546-1718
1546-1718
DOI:10.1038/s41588-018-0257-y