Advanced search

Extreme value theory for threshold determination in droplet digital PCR experiments

Matthijs Vynck (UGent) , Wim Trypsteen (UGent) , Jan De Neve (UGent) , Pawel Bonczkowski (UGent) , Maja Kiselinova, Eva Malatinková (UGent) , KAREN VERVISCH (UGent) , Olivier Thas (UGent) , Linos Vandekerckhove (UGent) and Ward De Spiegelaere (UGent)
Author
Organization
Abstract
Droplet digital PCR (ddPCR) technology can be used to quantify nucleic acid concentrations by measuring the fluorescence intensity for each single droplet and subsequently calling the droplet as positive or negative depending on its fluorescence intensity. Thus, setting a correct fluorescence level threshold is needed to accurately identify the presence (absence) of target nucleic acid. Currently available methods assume a normal distribution of the fluorescence readout. We show that this assumption does not hold in a large majority of assessed experiments and that wrongfully relying on this assumption typically results in an increase of false positives. By relying on extreme value theory (EVT) we do not make any assumptions on the droplet fluorescence intensity distribution. Using this EVT framework we develop a methodology for modeling the maximum fluorescence intensity of the negative droplets using negative control samples (no target nucleic acid present). We show how we subsequently select a final threshold. We discuss the (dis)advantages of relying on EVT and detail some important considerations when applying EVT and how this translates to the ddPCR case.

Citation

Please use this url to cite or link to this publication:

Chicago
Vynck, Matthijs, Wim Trypsteen, Jan De Neve, Pawel Bonczkowski, Maja Kiselinova, Eva Malatinková, KAREN VERVISCH, Olivier Thas, Linos Vandekerckhove, and Ward De Spiegelaere. 2015. “Extreme Value Theory for Threshold Determination in Droplet Digital PCR Experiments.” In Advances in qPCR & dPCR, Abstracts.
APA
Vynck, M., Trypsteen, W., De Neve, J., Bonczkowski, P., Kiselinova, M., Malatinková, E., VERVISCH, K., et al. (2015). Extreme value theory for threshold determination in droplet digital PCR experiments. Advances in qPCR & dPCR, Abstracts. Presented at the 6th Annual conference on Advances in qPCR & dPCR.
Vancouver
1.
Vynck M, Trypsteen W, De Neve J, Bonczkowski P, Kiselinova M, Malatinková E, et al. Extreme value theory for threshold determination in droplet digital PCR experiments. Advances in qPCR & dPCR, Abstracts. 2015.
MLA
Vynck, Matthijs, Wim Trypsteen, Jan De Neve, et al. “Extreme Value Theory for Threshold Determination in Droplet Digital PCR Experiments.” Advances in qPCR & dPCR, Abstracts. 2015. Print.
@inproceedings{5992806,
  abstract     = {Droplet digital PCR (ddPCR) technology can be used to quantify nucleic acid concentrations by measuring the fluorescence intensity for each single droplet and subsequently calling the droplet as positive or negative depending on its fluorescence intensity. Thus, setting a correct fluorescence level threshold is needed to accurately identify the presence (absence) of target nucleic acid.
Currently available methods assume a normal distribution of the fluorescence readout. We show that this assumption does not hold in a large majority of assessed experiments and that wrongfully relying on this assumption typically results in an increase of false positives.
By relying on extreme value theory (EVT) we do not make any assumptions on the droplet fluorescence intensity distribution. Using this EVT framework we develop a methodology for modeling the maximum fluorescence intensity of the negative droplets using negative control samples (no target nucleic acid present). We show how we subsequently select a final threshold. We discuss the (dis)advantages of relying on EVT and detail some important considerations when applying EVT and how this translates to the ddPCR case.},
  author       = {Vynck, Matthijs and Trypsteen, Wim and De Neve, Jan and Bonczkowski, Pawel and Kiselinova, Maja and Malatinková, Eva and VERVISCH, KAREN and Thas, Olivier and Vandekerckhove, Linos and De Spiegelaere, Ward},
  booktitle    = {Advances in qPCR & dPCR, Abstracts},
  language     = {eng},
  location     = {Singapore, Singapore},
  title        = {Extreme value theory for threshold determination in droplet digital PCR experiments},
  year         = {2015},
}