EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Developing a Machine Learning ‘Smart’ Polymerase Chain Reaction Thermocycler Part 2: Putting the Theoretical Framework into Practice

Caitlin McDonald iD, Duncan Taylor iD, Russell S A Brinkworth iD, Adrian Linacre

Genes · 2024

DOI: 10.3390/genes15091199 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Genes, 2024. Verfasst von Caitlin McDonald und 3 weiteren Autor:innen. Beteiligte Einrichtung: College of Science and Engineering, Flinders University, GPO Box 2100, Adelaide, SA 5001, Australia. Hauptthemen laut MEDLINE: Machine Learning, Polymerase Chain Reaction. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 27 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Attorney General’s Department, Forensic Science SA, Ross Vining Research.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The introduction of PCR into forensic science and the rapid increases in the sensitivity, specificity and discrimination power of DNA profiling that followed have been fundamental in shaping the field of forensic biology. Despite these developments, the challenges associated with the DNA profiling of trace, inhibited and degraded samples remain. Thus, any improvement to the performance of sub-optimal samples in DNA profiling would be of great value to the forensic community. The potential exists to optimise the PCR performance of samples by altering the cycling conditions used. If the effects of changing cycling conditions upon the quality of a DNA profile can be well understood, then the PCR process can be manipulated to achieve a specific goal. This work is a proof-of-concept study for the development of a smart PCR system, the theoretical foundations of which are outlined in part 1 of this publication. The first steps needed to demonstrate the performance of our smart PCR goal involved the manual alteration of cycling conditions and assessment of the DNA profiles produced. In this study, the timing and temperature of the denaturation and annealing stages of the PCR were manually altered to achieve the goal of reducing PCR runtime while maintaining an acceptable quality and quantity of DNA product. A real-time feedback system was also trialled using an STR PCR and qPCR reaction mix, and the DNA profiles generated were compared to profiles produced using the standard STR PCR kits. The aim of this work was to leverage machine learning to enable real-time adjustments during a PCR, allowing optimisation of cycling conditions towards predefined user goals. A set of parameters was found that yielded similar results to the standard endpoint PCR methodology but was completed 30 min faster. The development of an intelligent system would have significant implications for the various biological disciplines that are reliant on PCR technology.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Machine LearningPolymerase Chain ReactionHumansDNA FingerprintingDNAForensic Genetics

Stichwörter der Autor:innen

PCR thermocyclerSTR DNA profilecycling conditionsmachine learning

Substanzen

DNA

Autor:innen und Institutionen

  • College of Science and Engineering, Flinders University, GPO Box 2100, Adelaide, SA 5001, Australia.
  • Forensic Science SA, GPO Box 2790, Adelaide, SA 5001, Australia.

ORCID-Profile: Caitlin McDonald, Duncan Taylor, Russell S A Brinkworth

Förderung

  • Attorney General’s Department
  • Forensic Science SA
  • Ross Vining Research

Zitieren

APA

McDonald, C., Taylor, D., Brinkworth, R. S. A., & Linacre, A. (2024). Developing a Machine Learning ‘Smart’ Polymerase Chain Reaction Thermocycler Part 2: Putting the Theoretical Framework into Practice. Genes. https://doi.org/10.3390/genes15091199

Vancouver

McDonald C, Taylor D, Brinkworth RSA, Linacre A. Developing a Machine Learning ‘Smart’ Polymerase Chain Reaction Thermocycler Part 2: Putting the Theoretical Framework into Practice. Genes. 2024. doi:10.3390/genes15091199

BibTeX
@article{McDonald2024,
  author = {McDonald, Caitlin and Taylor, Duncan and Brinkworth, Russell S A and Linacre, Adrian},
  title = {Developing a Machine Learning ‘Smart’ Polymerase Chain Reaction Thermocycler Part 2: Putting the Theoretical Framework into Practice},
  journal = {Genes},
  year = {2024},
  doi = {10.3390/genes15091199}
}
Kostenlos starten →