Comparative performances of machine learning methods for classifying Crohn Disease patients using genome-wide genotyping data.
View/ Open
Publication Date
2019-07-17ICR Author
Author
Romagnoni, A
Jégou, S
Van Steen, K
Wainrib, G
Hugot, J-P
International Inflammatory Bowel Disease Genetics Consortium (IIBDGC)
Type
Journal Article
Metadata
Show full item recordAbstract
Crohn Disease (CD) is a complex genetic disorder for which more than 140 genes have been identified using genome wide association studies (GWAS). However, the genetic architecture of the trait remains largely unknown. The recent development of machine learning (ML) approaches incited us to apply them to classify healthy and diseased people according to their genomic information. The Immunochip dataset containing 18,227 CD patients and 34,050 healthy controls enrolled and genotyped by the international Inflammatory Bowel Disease genetic consortium (IIBDGC) has been re-analyzed using a set of ML methods: penalized logistic regression (LR), gradient boosted trees (GBT) and artificial neural networks (NN). The main score used to compare the methods was the Area Under the ROC Curve (AUC) statistics. The impact of quality control (QC), imputing and coding methods on LR results showed that QC methods and imputation of missing genotypes may artificially increase the scores. At the opposite, neither the patient/control ratio nor marker preselection or coding strategies significantly affected the results. LR methods, including Lasso, Ridge and ElasticNet provided similar results with a maximum AUC of 0.80. GBT methods like XGBoost, LightGBM and CatBoost, together with dense NN with one or more hidden layers, provided similar AUC values, suggesting limited epistatic effects in the genetic architecture of the trait. ML methods detected near all the genetic variants previously identified by GWAS among the best predictors plus additional predictors with lower effects. The robustness and complementarity of the different methods are also studied. Compared to LR, non-linear models such as GBT or NN may provide robust complementary approaches to identify and classify genetic markers.
Collections
Licenseref URL
http://creativecommons.org/licenses/by/4.0/Version of record
Subject
International Inflammatory Bowel Disease Genetics Consortium (IIBDGC)
Humans
Crohn Disease
Genetic Predisposition to Disease
Area Under Curve
Logistic Models
ROC Curve
Genotype
Polymorphism, Single Nucleotide
Alleles
Decision Trees
Nonlinear Dynamics
Models, Genetic
Female
Male
INDEL Mutation
Genome-Wide Association Study
Genotyping Techniques
Deep Learning
Neural Networks, Computer
Language
eng
Date accepted
2019-07-03
License start date
2019-07-17
Citation
Scientific reports, 2019, 9 (1), pp. 10351 - ?
Except where otherwise noted, this item's license is described as http://creativecommons.org/licenses/by/4.0/
Related items
Showing items related by title, author, creator and subject.
-
EFSUMB guidelines and recommendations on the clinical use of ultrasound elastography. Part 2: Clinical applications.
Cosgrove, D; Piscaglia, F; Bamber, J; Bojunga, J; Correas, J-M; Gilja, OH; Klauser, AS; Sporea, I; Calliada, F; Calliada, F; Cantisani, V; D'Onofrio, M; Drakonaki, EE; Fink, M; Friedrich-Rust, M; Fromageau, J; Havre, RF; Jenssen, C; Ohlinger, R; Săftoiu, A; Schaefer, F; Dietrich, CF; EFSUMB (2013-06)The clinical part of these Guidelines and Recommendations produced under the auspices of the European Federation of Societies for Ultrasound in Medicine and Biology EFSUMB assesses the clinically used applications of all ... -
Enzalutamide in castration-resistant prostate cancer patients with visceral disease in the liver and/or lung: Outcomes from the randomized controlled phase 3 AFFIRM trial.
Loriot, Y; Fizazi, K; de Bono, JS; Forer, D; Hirmand, M; Scher, HI (2017-01)<h4>Background</h4>Patients with metastatic castration-resistant prostate cancer (mCRPC) and visceral metastases have a worse prognosis than those with nonvisceral metastases. Treatment with the androgen receptor inhibitor ... -
Trial Design and Objectives for Castration-Resistant Prostate Cancer: Updated Recommendations From the Prostate Cancer Clinical Trials Working Group 3.
Scher, HI; Morris, MJ; Stadler, WM; Higano, C; Basch, E; Fizazi, K; Antonarakis, ES; Beer, TM; Carducci, MA; Chi, KN; Corn, PG; de Bono, JS; Dreicer, R; George, DJ; Heath, EI; Hussain, M; Kelly, WK; Liu, G; Logothetis, C; Nanus, D; Stein, MN; Rathkopf, DE; Slovin, SF; Ryan, CJ; Sartor, O; Small, EJ; Smith, MR; Sternberg, CN; Taplin, M-E; Wilding, G; Nelson, PS; Schwartz, LH; Halabi, S; Kantoff, PW; Armstrong, AJ; Prostate Cancer Clinical Trials Working Group 3 (2016-04)<h4>Purpose</h4>Evolving treatments, disease phenotypes, and biology, together with a changing drug development environment, have created the need to revise castration-resistant prostate cancer (CRPC) clinical trial ...