Atif Khan1, Thomas H Thatcher, Collynn F Woeller, Patricia J Sime, Richard P Phipps, Philip K Hopke, Mark J Utell, Pamela L Krahl, Timothy M Mallon, Juilee Thakar. 1. Departments of Microbiology and Immunology and Biostatistics and Computational Biology (Mr Khan, Ms Thakar); Department of Environmental Medicine (Mr Woeller); Departments of Medicine, Environmental Medicine, and Microbiology and Immunology (Mr Phipps); Department of Public Health Sciences (Mr Hopke); Center for Air Resources Engineering and Science, Clarkson University, Potsdam (Mr Hopke); Departments of Medicine and Environmental Medicine (Mr Utell), University of Rochester Medical Center, Rochester, New York; Division of Pulmonary Diseases and Critical Care Medicine, Virginia Commonwealth University, Richmond, Virginia (Mr Thatcher, Ms Sime); Department of Preventive Medicine and Biostatistics, Uniformed Services University of the Health Sciences, Bethesda, Maryland (Ms Krahl, Mr Mallon).
Abstract
OBJECTIVE: To develop an approach for a retrospective analysis of post-exposure serum samples using diverse molecular profiles. METHODS: The 236 molecular profiles from 800 de-identified human serum samples from the Department of Defense Serum Repository were classified as smokers or non-smokers based on direct measurement of serum cotinine levels. A machine-learning pipeline was used to classify smokers and non-smokers from their molecular profiles. RESULTS: The refined supervised support vector machines with recursive feature elimination predicted smokers and non-smokers with 78% accuracy on the independent held-out set. Several of the identified classifiers of smoking status have previously been reported and four additional miRNAs were validated with experimental tobacco smoke exposure in mice, supporting the computational approach. CONCLUSIONS: We developed and validated a pipeline that shows retrospective analysis of post-exposure serum samples can identify environmental exposures.
OBJECTIVE: To develop an approach for a retrospective analysis of post-exposure serum samples using diverse molecular profiles. METHODS: The 236 molecular profiles from 800 de-identified human serum samples from the Department of Defense Serum Repository were classified as smokers or non-smokers based on direct measurement of serum cotinine levels. A machine-learning pipeline was used to classify smokers and non-smokers from their molecular profiles. RESULTS: The refined supervised support vector machines with recursive feature elimination predicted smokers and non-smokers with 78% accuracy on the independent held-out set. Several of the identified classifiers of smoking status have previously been reported and four additional miRNAs were validated with experimental tobacco smoke exposure in mice, supporting the computational approach. CONCLUSIONS: We developed and validated a pipeline that shows retrospective analysis of post-exposure serum samples can identify environmental exposures.
Authors: Salud Santos; Victor I Peinado; Josep Ramirez; Jaime Morales-Blanhir; Ricardo Bastos; Josep Roca; Robert Rodriguez-Roisin; Joan A Barbera Journal: Am J Respir Crit Care Med Date: 2003-02-20 Impact factor: 21.405
Authors: Tariq A Bhat; Suresh Gopi Kalathil; Paul N Bogner; Austin Miller; Paul V Lehmann; Thomas H Thatcher; Richard P Phipps; Patricia J Sime; Yasmin Thanavala Journal: J Immunol Date: 2018-03-19 Impact factor: 5.422
Authors: Neal L Benowitz; John T Bernert; Ralph S Caraballo; David B Holiday; Jiantong Wang Journal: Am J Epidemiol Date: 2008-11-19 Impact factor: 4.897
Authors: Jason Liu; Nicholas Lezama; Joseph Gasper; Jennifer Kawata; Sybil Morley; Drew Helmer; Paul Ciminera Journal: J Occup Environ Med Date: 2016-07 Impact factor: 2.162
Authors: Xuegong Zhang; Xin Lu; Qian Shi; Xiu-Qin Xu; Hon-Chiu E Leung; Lyndsay N Harris; James D Iglehart; Alexander Miron; Jun S Liu; Wing H Wong Journal: BMC Bioinformatics Date: 2006-04-10 Impact factor: 3.169
Authors: Hsi-Min Hsiao; Ramil E Sapinoro; Thomas H Thatcher; Amanda Croasdell; Elizabeth P Levy; Robert A Fulton; Keith C Olsen; Stephen J Pollock; Charles N Serhan; Richard P Phipps; Patricia J Sime Journal: PLoS One Date: 2013-03-06 Impact factor: 3.240