TargetCall: Eliminating the Wasted Computation in Basecalling via Pre-Basecalling Filtering

Abstract

Basecalling is an essential step in nanopore sequencing analysis where the raw signals of nanopore sequencers are converted into nucleotide sequences, i.e., reads. State-of-the-art basecallers employ complex deep learning models to achieve high basecalling accuracy. This makes basecalling computationally-inefficient and memory-hungry; bottlenecking the entire genome analysis pipeline. However, for many applications, the majority of reads do no match the reference genome of interest (i.e., target reference) and thus are discarded in later steps in the genomics pipeline, wasting the basecalling computation. To overcome this issue, we propose TargetCall, the first fast and widely-applicable pre-basecalling filter to eliminate the wasted computation in basecalling. TargetCall's key idea is to discard reads that will not match the target reference (i.e., off-target reads) prior to basecalling. TargetCall consists of two main components: (1) LightCall, a lightweight neural network basecaller that produces noisy reads; and (2) Similarity Check, which labels each of these noisy reads as on-target or off-target by matching them to the target reference. TargetCall filters out all off-target reads before basecalling; and the highly-accurate but slow basecalling is performed only on the raw signals whose noisy reads are labeled as on-target. Our thorough experimental evaluations using both real and simulated data show that TargetCall 1) improves the end-to-end basecalling performance of the state-of-the-art basecaller by 3.31x while maintaining high (98.88%) sensitivity in keeping on-target reads, 2) maintains high accuracy in downstream analysis, 3) precisely filters out up to 94.71% of off-target reads, and 4) achieves better performance, sensitivity, and generality compared to prior works. We freely open-source TargetCall to aid future research in pre-basecalling filtering at https://github.com/CMU-SAFARI/TargetCall.

RawHash: Enabling Fast and Accurate Real-Time Analysis of Raw Nanopore Signals for Large Genomes

“…Although we do not evaluate in this work, we expect that RawHash can be used as a low-cost filter to eliminate the reads that are unlikely to be useful in downstream analysis, which can reduce the overall workload of basecallers and further downstream analysis. We believe that RawHash can be applied for such a filtering purpose since a previous work [39] proposes a lightweight basecalling solution to use as a filter before using the costly basecallers. Limitations.…”

“…All the above works accelerate the basecalling step without eliminating the wasted computation in basecalling. TargetCall [39] proposes a pre-basecalling filter that eliminates the wasted computation in basecalling by leveraging the observation that the majority of reads are discarded after basecalling. However, RawHash is different from these works as its goal is to perform real-time analysis of raw signals without performing the computationally-intensive basecalling step.…”

Sigmoni: classification of nanopore signal with a compressed pangenome index

“…Previous studies have used lightweight basecallers for faster adaptive sampling [13]. Our picoamp binning procedure could be considered a very simple "basecalling" step.…”

“…However, these networks comprise millions of parameters and are computationally infeasible to run without hardware acceleration [12]. Even using acceleration, basecalling is the rate-limiting step for read classification algorithms [13]. One study found that over 95% of the compute time in a variant calling pipeline for SARS-CoV2 was spent in basecalling [14].…”

“…SPUMONI 2 also introduces a document array data structure for multi-class classification. TargetCall [13] introduced the idea of pre-basecalling filtering, using simpler neural-network basecallers to identify reads of interest, which are then re-basecalled with larger, slower models. All of these methods rely on basecalling, and therefore do not address the main bottleneck.…”