[10p-N207-1~16] 12.6 Nanobiotechnology

[10p-N207-10] Deep learning enhanced single-particle motion tracking in solid-state nanopores

〇Makusu Tsutsui¹, Takayuki Takaai¹, Kazumichi Yokota², Tomoji Kawai¹, Takashi Washio¹ (1.Osaka Univ., 2.AIST)

Keywords:nanopore, deep learning, ionic current

Silid-state nanopore is a versatile platform for sensing nanoscale objects in liquid. A long-standing issue in this sensor has been the fast translocation motions of analytes that makes it difficult to detect meaningful information from resistive pulse signals due to relatively large noise floor involved in the ionic current measurements. In the present study, therefore, we employed deep neural networks to denoise the high-frequency noise in a high dimensional feature space. It enabled to identify particle-motion-derived faint changes in the ionic current hidden in the large noise.