Nanopore Base Calling on the Edge

TL;DR

DeepNano-coral is a novel, energy-efficient nanopore base caller optimized for edge hardware, achieving real-time sequencing with competitive accuracy by designing new neural network components that reduce memory access.

Contribution

We introduce DeepNano-coral, a base caller optimized for edge devices, with new neural network components that improve efficiency and enable real-time nanopore sequencing.

Findings

Achieves real-time base calling on Coral Edge TPU.

Slightly better accuracy than Guppy's fast mode.

Consumes only 10W of power, demonstrating high energy efficiency.

Abstract

We developed a new base caller DeepNano-coral for nanopore sequencing, which is optimized to run on the Coral Edge Tensor Processing Unit, a small USB-attached hardware accelerator. To achieve this goal, we have designed new versions of two key components used in convolutional neural networks for speech recognition and base calling. In our components, we propose a new way of factorization of a full convolution into smaller operations, which decreases memory access operations, memory access being a bottleneck on this device. DeepNano-coral achieves real-time base calling during sequencing with the accuracy slightly better than the fast mode of the Guppy base caller and is extremely energy efficient, using only 10W of power. Availability: https://github.com/fmfi-compbio/coral-basecaller