Title

The identification of menstrual blood in forensic samples by logistic regression modeling of miRNA expression

Authors

Authors

E. K. Hanson; M. Mirza; K. Rekab;J. Ballantyne

Comments

Authors: contact us about adding a copy of your work at STARS@ucf.edu

Abbreviated Journal Title

Electrophoresis

Keywords

Body fluid identification; Forensic science; Logistic regression; analysis; MicroRNA (miRNA); RNA profiling; BODY-FLUID IDENTIFICATION; MESSENGER-RNA MARKERS; COLLABORATIVE EDNAP; EXERCISE; RNA/DNA CO-ANALYSIS; QUANTITATIVE RT-PCR; REAL-TIME PCR; DNA; METHYLATION; CAPILLARY-ELECTROPHORESIS; REVERSE TRANSCRIPTION; MICRORNA; MARKERS; Biochemical Research Methods; Chemistry, Analytical

Abstract

We report the identification of sensitive and specific miRNA biomarkers for menstrual blood, a tissue that might provide probative information in certain specialized instances. We incorporated these biomarkers into qPCR assays and developed a quantitative statistical model using logistic regression that permits the prediction of menstrual blood in a forensic sample with a high, and measurable, degree of accuracy. Using the developed model, we achieved 100% accuracy in determining the body fluid of interest for a set of test samples (i.e. samples not used in model development). The development, and details, of the logistic regression model are described. Testing and evaluation of the finalized logistic regression modeled assay using a small number of samples was carried out to preliminarily estimate the limit of detection (LOD), specificity in admixed samples and expression of the menstrual blood miRNA biomarkers throughout the menstrual cycle (25-28 days). The LOD was <1 ng of total RNA, the assay performed as expected with admixed samples and menstrual blood was identified only during the menses phase of the female reproductive cycle in two donors.

Journal Title

Electrophoresis

Volume

35

Issue/Number

21-22

Publication Date

1-1-2014

Document Type

Article

Language

English

First Page

3087

Last Page

3095

WOS Identifier

WOS:000345272200010

ISSN

0173-0835

Share

COinS