Related Experiment Video
Updated: Jul 13, 2026
13:11
Optimized Analysis of DNA Methylation and Gene Expression from Small, Anatomically-defined Areas of the Brain
Published on: July 12, 2012
19.5K
DNA methylation-based assay and Random Forest classification model for identification of biological materials
Laura Schmelzer1, Franz Haehn1, Ulrike Schmidt1
1Institute of Forensic Medicine, Medical Center - University of Freiburg, Medical Faculty - University of Freiburg, Germany.
Forensic Science International. Genetics
|April 7, 2026
Summary
This study introduces a new DNA methylation assay to identify biological materials like semen and blood in forensic investigations. The assay shows high accuracy for single samples and potential for analyzing complex mixtures.
Area of Science:
- Forensic Science
- Molecular Biology
- Genetics
Background:
- Determining the biological origin of trace evidence is crucial in forensic investigations.
- Differentially methylated DNA positions (DMPs) offer a method to distinguish between various biological materials.
Purpose of the Study:
- To develop and validate a multiplex assay targeting 13 DMPs for identifying forensically relevant biological materials.
- To build predictive models for classifying single-component samples and mixtures using DNA methylation profiles.
Main Methods:
- A multiplex assay using combined bisulfite conversion and SNaPshot analysis was developed.
- Random Forest models (RF1 for single samples, RF2 for mixtures) were trained and validated.
- DNA methylation profiles were analyzed in semen, blood, saliva, skin, and other samples, including in-silico mixtures.
Main Results:
- The assay achieved 92.7% accuracy in identifying single-component samples.
- DNA methylation levels showed a strong linear correlation with mixture component proportions.
- The in-silico mixture model demonstrated potential for identifying unbalanced experimental mixtures.
Conclusions:
- The developed DMP assay is a valuable tool for forensic identification of biological materials.
- The Random Forest models show promise for practical application in analyzing both single samples and complex mixtures.
- Further research can refine the assay's performance with low DNA input and diverse sample types.
