Napel

Singh, Gagandeep; Gómez-Luna, Juan; Mariani, Giovanni; Oliveira, Geraldo F.; Corda, Stefano; Stuijk, Sander; Mutlu, Onur; Corporaal, Henk

doi:10.1145/3316781.3317867

Cited by 62 publications

(2 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, accelerates the basecaller architecture search to develop high-performance basecalling architectures. The final model architecture can be further fine-tuned for other hyperparameters [80,81], such as learning rate and batch size (for example, with grid search or neural architecture search). Throughout our experiments, we build general-purpose basecalling models by training and testing the model using an official, open-source ONT dataset that consists of a mix of different species.…”

Section: Discussionmentioning

confidence: 99%

A Framework for Designing Efficient Deep Learning-Based Genomic Basecallers

Singh

Alser

Khodamoradi³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Nanopore sequencing is a widely-used high-throughput genome sequencing technology that can sequence long fragments of a genome. Nanopore sequencing generates noisy electrical signals that need to be converted into a standard string of DNA nucleotide bases (i.e., A, C, G, T) using a computational step called basecalling. The accuracy and speed of basecalling have critical implications for every subsequent step in genome analysis. Currently, basecallers are mainly based on deep learning techniques to provide high sequencing accuracy without considering the compute demands of such tools. We observe that state-of-the-art basecallers (i.e., Guppy, Bonito) are slow, inefficient, and memory-hungry as researchers have adapted deep learning models from other domains without specialization to the basecalling purpose. Our goal is to make basecalling highly efficient and fast by building the first framework for specializing and optimizing machine learning-based basecaller. We introduce RUBICON, a framework to develop hardware-optimized basecallers. RUBICON consists of two novel machine-learning techniques that are specifically designed for basecalling. First, we introduce the quantization-aware basecalling neural architecture search (QABAS) framework to specialize the basecalling neural network architecture for a given hardware acceleration platform while jointly exploring and finding the best bit-width precision for each neural network layer. Second, we develop SkipClip, the first technique to remove all the skip connections present in modern basecallers to greatly reduce resource and storage requirements without any loss in basecalling accuracy. We demonstrate the benefits of QABAS and SkipClip by developing RUBICALL, the first hardware-optimized basecaller that performs fast and accurate basecalling. Our experimental results on state-of-the-art computing systems show that RUBICALL is a fast, accurate and hardware-friendly, mixed-precision basecaller. Compared to a highly-accurate state-of-the-art basecaller, RUBICALL provides a 16.56x speedup without losing accuracy, while also achieving a 6.88x and 2.94x reduction in neural network model size and the number of parameters, respectively. Compared to the fastest state-of-the-art basecaller, RUBICALL provides a 3.19x speedup with 2.97% higher accuracy. We show that QABAS and SkipClip can help researchers develop hardware-optimized basecallers that are superior to expert-designed models.

show abstract

Section: Discussionmentioning

confidence: 99%

A Framework for Designing Efficient Deep Learning-Based Genomic Basecallers

Singh

Alser

Khodamoradi³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Moreover, in [19], Random Forest was used for HLS design space exploration. Similarly, in [20], Random Forest regression is used for predicting instruction per cycle for near-memory computing applications. In [21] Markov decision process was found effective for the design space exploration of multi-processor platforms.…”

Section: Related Workmentioning

confidence: 99%