Abstract

We present SquiggleNet, the first deep-learning model that can classify nanopore reads directly from their electrical signals. SquiggleNet operates faster than DNA passes through the pore, allowing real-time classification and read ejection. Using 1 s of sequencing data, the classifier achieves significantly higher accuracy than base calling followed by sequence alignment. Our approach is also faster and requires an order of magnitude less memory than alignment-based approaches. SquiggleNet distinguished human from bacterial DNA with over 90% accuracy, generalized to unseen bacterial species in a human respiratory meta genome sample, and accurately classified sequences containing human long interspersed repeat elements.

Highlights

  • Oxford Nanopore sequencers, such as MinION or PromethION, determine the nucleotide sequence of a DNA or RNA molecule by measuring changes in electrical current as the molecule translocates through a protein nanopore

  • The changes in electrical current induced by a DNA or RNA molecule depend on the specific chemical properties of the nucleotides, including secondary structure interactions and epigenomic modifications such as methylation

  • SquiggleNet: a convolutional neural network for classifying nanopore signals SquiggleNet is a deep neural network that classifies molecules of interest based on statistical patterns in nanopore conductivity, which are often hard for humans to identify by eye, automatically extracted from the input data

Read more

Summary

Introduction

Oxford Nanopore sequencers, such as MinION or PromethION, determine the nucleotide sequence of a DNA or RNA molecule by measuring changes in electrical current (called “squiggles”) as the molecule translocates through a protein nanopore This approach is fundamentally different from the widely-used Illumina platform and provides several benefits: the MinION is small, fast, and portable, making it ideal for rapid diagnostics and field work. Because it does not rely upon synchronized nucleotide addition (the heart of the Illumina sequencing-by-synthesis technology), MinION produces much longer reads. The nanopore sequencer can stream the squiggle data to a computer in real time

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call