Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Artificial Intelligence Approach for Variant Reporting

Artificial Intelligence Approach for Variant Reporting Purpose: Next-generation sequencing technologies are actively applied in clinical oncology. Bioinformatics pipeline analysis is an integral part of this process; however, humans cannot yet realize the full potential of the highly complex pipeline output. As a result, the decision to include a variant in the final report during routine clinical sign-out remains challenging. Methods: We used an artificial intelligence approach to capture the collective clinical sign-out experience of six board-certified molecular pathologists to build and validate a decision support tool for variant reporting. We extracted all reviewed and reported variants from our clinical database and tested several machine learning models. We used 10-fold cross-validation for our variant call prediction model, which derives a contiguous prediction score from 0 to 1 (no to yes) for clinical reporting. Results: For each of the 19,594 initial training variants, our pipeline generates approximately 500 features, which results in a matrix of > 9 million data points. From a comparison of naive Bayes, decision trees, random forests, and logistic regression models, we selected models that allow human interpretability of the prediction score. The logistic regression model demonstrated 1% false negativity and 2% false positivity. The final models' Youden indices were 0.87 and 0.77 for screening and confirmatory cutoffs, respectively. Retraining on a new assay and performance assessment in 16,123 independent variants validated our approach (Youden index, 0.93). We also derived individual pathologist-centric models (virtual consensus conference function), and a visual drill-down functionality allows assessment of how underlying features contributed to a particular score or decision branch for clinical implementation. Conclusion: Our decision support tool for variant reporting is a practically relevant artificial intelligence approach to harness the next-generation sequencing bioinformatics pipeline output when the complexity of data interpretation exceeds human capabilities. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JCO Clinical Cancer Informatics Wolters Kluwer Health

Artificial Intelligence Approach for Variant Reporting

Download PDF
13 pages

Loading next page...
 
/lp/wolters-kluwer-health/artificial-intelligence-approach-for-variant-reporting-AgMfjqJNML

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Wolters Kluwer Health
Copyright
(C) 2018 by Lippincott Williams & Wilkins, Inc.
ISSN
2473-4276
DOI
10.1200/CCI.16.00079
Publisher site
See Article on Publisher Site

Abstract

Purpose: Next-generation sequencing technologies are actively applied in clinical oncology. Bioinformatics pipeline analysis is an integral part of this process; however, humans cannot yet realize the full potential of the highly complex pipeline output. As a result, the decision to include a variant in the final report during routine clinical sign-out remains challenging. Methods: We used an artificial intelligence approach to capture the collective clinical sign-out experience of six board-certified molecular pathologists to build and validate a decision support tool for variant reporting. We extracted all reviewed and reported variants from our clinical database and tested several machine learning models. We used 10-fold cross-validation for our variant call prediction model, which derives a contiguous prediction score from 0 to 1 (no to yes) for clinical reporting. Results: For each of the 19,594 initial training variants, our pipeline generates approximately 500 features, which results in a matrix of > 9 million data points. From a comparison of naive Bayes, decision trees, random forests, and logistic regression models, we selected models that allow human interpretability of the prediction score. The logistic regression model demonstrated 1% false negativity and 2% false positivity. The final models' Youden indices were 0.87 and 0.77 for screening and confirmatory cutoffs, respectively. Retraining on a new assay and performance assessment in 16,123 independent variants validated our approach (Youden index, 0.93). We also derived individual pathologist-centric models (virtual consensus conference function), and a visual drill-down functionality allows assessment of how underlying features contributed to a particular score or decision branch for clinical implementation. Conclusion: Our decision support tool for variant reporting is a practically relevant artificial intelligence approach to harness the next-generation sequencing bioinformatics pipeline output when the complexity of data interpretation exceeds human capabilities.

Journal

JCO Clinical Cancer Informatics – Wolters Kluwer Health

Published: Mar 22, 2018

References

  • The evolving war on cancer
    Haber DA; Gray NS; Baselga J
  • The evolution of molecular genetic pathology: Advancing 20th-century diagnostic methods into potent tools for the new millennium
    Sobel ME; Bagg A; Caliendo AM
  • Coming of age: Ten years of next-generation sequencing technologies
    Goodwin S; McPherson JD; McCombie WR
  • Next generation sequencing technology: Advances and applications
    Buermans HP; den Dunnen JT
  • Diagnostic yield of targeted next generation sequencing in various cancer types: An information-theoretic approach
    Hagemann IS; O'Neill PK; Erill I
  • Next-generation sequencing informatics: Challenges and strategies for implementation in a clinical environment
    Roy S; LaFramboise WA; Nikiforov YE
  • Mutational landscape of metastatic cancer revealed from prospective clinical sequencing of 10,000 patients
    Zehir A; Benayed R; Shah RH
  • Clinical genomicist workstation
    Sharma MK; Phillips J; Agarwal S
  • Standards for clinical grade genomic databases
    Yohe SL; Carter AB; Pfeifer JD
  • Precision diagnosis is a team sport
    Zehnbauer BA; Buchman TG
  • College of American Pathologists' laboratory standards for next-generation sequencing clinical tests
    Aziz N; Zhao Q; Bry L
  • VarDict: A novel and versatile variant caller for next-generation sequencing in cancer research
    Lai Z; Markovets A; Ahdesmaki M
  • Variant callers for next-generation sequencing data: A comparison study
    Liu X; Han S; Wang Z
  • A general framework for estimating the relative pathogenicity of human genetic variants
    Kircher M; Witten DM; Jain P
  • Evaluation of nine somatic variant callers for detection of somatic mutations in exome and targeted deep sequencing data
    Kroigard AB; Thomassen M; Laenkholm AV
  • Standardized decision support in next generation sequencing reports of somatic cancer variants
    Dienstmann R; Dong F; Borger D
  • The cancer genomics resource list 2014
    Zutter MM; Bloom KJ; Cheng L
  • Enhancing next-generation sequencing-guided cancer care through cognitive computing
    Patel NM; Michelini VV; Snell JM
  • Clinical decision support systems for improving diagnostic accuracy and achieving precision medicine
    Castaneda C; Nalley K; Mannion C
  • The 2013 symposium on pathology data integration and clinical decision support and the current state of field
    Baron JM; Dighe AS; Arnaout R
  • The scientists' apprentice
    Appenzeller T
  • AI glossary: Artificial intelligence, in so many words
    Hutson M
  • Artificial intelligence in research
    Musib M; Wang F; Tarselli MA
  • Perspectives on making big data analytics work for oncology
    El Naqa I
  • The rise of artificial intelligence and the uncertain future for physicians
    Krittanawong C
  • Towards automatic pulmonary nodule management in lung cancer screening with deep learning
    Ciompi F; Chung K; van Riel SJ
  • DeepStack: Expert-level artificial intelligence in heads-up no-limit poker
    Moravcik M; Schmid M; Burch N
  • A method based on artificial intelligence to fully automatize the evaluation of bovine blastocyst images
    Rocha JC; Passalia FJ; Matos FD
  • Solving multiple isolated, interleaved, and blended tasks through modular neuroevolution
    Schrum J; Miikkulainen R
  • Systematic computational identification of variants that activate exonic and intronic cryptic splice sites
    Lee M; Roos P; Sharma N
  • Machine learning-assisted network inference approach to identify a new class of genes that coordinate the functionality of cancer networks
    Ghanat Bari M; Ung CY; Zhang C
  • Cognitive biases and decision support systems development: A design science approach
    Arnott D
  • Rapid targeted mutational analysis of human tumours: A clinical platform to guide personalized cancer medicine
    Dias-Santagata D; Akhavanfard S; David SS
  • Implementing multiplexed genotyping of non-small-cell lung cancers into routine clinical practice
    Sequist LV; Heist RS; Shaw AT
  • Anchored multiplex PCR for targeted next-generation sequencing
    Zheng Z; Liebers M; Zhelyazkova B
  • Fast and accurate short read alignment with Burrows-Wheeler transform
    Li H; Durbin R
  • Sensitive detection of somatic point mutations in impure and heterogeneous cancer samples
    Cibulskis K; Lawrence MS; Carter SL
  • Oncotator: Cancer variant annotation tool
    Ramos AH; Lichtenstein L; Gupta M
  • The Ensembl variant effect predictor
    McLaren W; Gil L; Hunt SE
  • A study of cross-validation and bootstrap for accuracy estimation and model selection
    Kohavi R
  • Machine Learning
    Kohavi R; Provost F
  • A survey of cross-validation procedures for model selection
    Arlot S; Celisse A
  • Sensitivity analysis of kappa-fold cross validation in prediction error estimation
    Rodriguez JD; Perez A; Lozano JA
  • A review of feature selection techniques in bioinformatics
    Saeys Y; Inza I; Larranaga P
  • Scikit-learn: Machine learning in Python
    Pedregosa F; Varoquax G; Michel V
  • Accurate and reproducible invasive breast cancer detection in whole-slide images: A deep learning approach for quantifying tumor extent
    Cruz-Roa A; Gilmore H; Basavanhally A
  • Prospective identification of hematopoietic lineage choice by deep learning
    Buggenthin F; Buettner F; Hoppe PS
  • Residual convolutional neural network for determination of IDH status in low- and high-grade gliomas from MR imaging
    Chang K; Bai HX; Zhou H
  • One hundred year study on artificial intelligence (AI100), 2016
    Stanford