Premium
Improving post‐traumatic stress disorder biomarkers through matching patients and controls on multiple criteria (1120.10)
Author(s) -
Donohue Duncan,
Miller StacyAnn,
Gupta Ellen,
Yang Ruoting,
Guatam Aarti,
Hammamieh Rasha,
JettTilton Marti
Publication year - 2014
Publication title -
the faseb journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.709
H-Index - 277
eISSN - 1530-6860
pISSN - 0892-6638
DOI - 10.1096/fasebj.28.1_supplement.1120.10
Subject(s) - biomarker , matching (statistics) , medicine , cohort , sampling (signal processing) , selection (genetic algorithm) , clinical psychology , psychology , computer science , biology , pathology , genetics , artificial intelligence , filter (signal processing) , computer vision
Molecular typing of psychological disease depends on a variety of complicated environmental and phenotypic factors. The quality of results critically depends on the cohort selection criteria, including those criteria used to match case and controls. Our PTSD biomarker study initially matched patient groups on age and self‐reported race. Further analysis has shown striking differences in the types and amounts of medications used by PTSD patients vs. controls. Here, we present an approach to simultaneously match patient groups across age, ancestry, and medication use applying random sampling methods. We compare receiver operating characteristics to quantify the degree to which matching on these three variables generates blood mRNA expression biomarkers which better discriminate patients from controls. Work is in progress to fine‐tune matching by incorporating the relative effect size of these three measures. Disclaimer: Research was conducted in compliance with the Animal Welfare Act and all other Federal requirements. The views expressed are those of the authors and do not constitute endorsement by the U.S. Army. Human blood sample use protocol was conducted at USACEHR under HRPO Log number A‐15459.f.