This dataset contains a set of nanopore translocation current traces. It is divided in two parts. Part I: This part contains artificially generated traces with different levels of background noise (SNR = 4, 2, 1, 0.5, and 0.25) For each noise level, three parameters are varied in data generation: a. Twenty different concentrations of nanoparticles as the analytes (Cnp): 0.013, 0.016, 0.020, 0.025, 0.032, 0.040, 0.050, 0.063, 0.080, 0.1, 0.13, 0.16, 0.20, 0.25, 0.32, 0.40, 0.50, 0.63, 0.80, and 1, with the unit of nano-molar, [nM]. b. Fifteen different diameters of the nanoparticles (Dnp): 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, and 17, with the unit of nanometer, [nm]. c. Five different translocation durations (Duration): 0.5, 1.0, 1.5, 3.0, and 5.0, with the unit of millisecond, [ms]. In total we have 20*15*5=1500 current traces for each SNR. There are three datasets: training, validation and test. Traces in training datasets are of 20 seconds, traces in validation and test datasets are 10 seconds long. For SNR = 4, training, validation and test datasets are provided. For other SNRs, only test traces are provided. Pert II: This part contains real experimental datasets, the translocation of Lambda DNA and Streptavidin with 6 current traces each at different bias voltages. Each Lambda DNA trace has 71 seconds, while each Streptavidin trace has 126 seconds. Two truncated pyramid shape nanopores were used in our experiments, one with a side length of 7.5 nm and another 16 nm, both in a 55 nm-thick silicon layer, for DNA and protein streptavidin translocation, respectively. The DNA and streptavidin were dispersed in 500 mM KCl electrolyte with a concentration of 78 pM and 84 nM, respectively. Details of artificially generated data and experimental data can be found in our paper: Dario Dematties, Chenyu Wen, Mauricio David Pérez, Dian Zhou, Shi-Li Zhang. Deep learning of nanopore sensing signals using a bi-path network. arXiv:2105.03660. Trained and Validated Models In this data set we also include all the trained and best validated models evaluated in our paper.

The Artificial data set is compressed and split into "n" .zip files. Once you downloaded such files you can use the following sequence of commands for joining and unzipping them: zip -F Artificial_Data.zip --out single-archive.zip unzip single-archive.zip