Lerna: transformer architectures for configuring error correction tools for short- and long-read genome sequencing

Abstract

AbstractBackgroundSequencing technologies are prone to errors, making error correction (EC) necessary for downstream applications. EC tools need to be manually configured for optimal performance. We find that the optimal parameters (e.g.,k-mer size) are both tool- and dataset-dependent. Moreover, evaluating the performance (i.e., Alignment-rate or Gain) of a given tool usually relies on a reference genome, but quality reference genomes are not always available. We introduceLernafor the automated configuration ofk-mer-based EC tools.Lernafirst creates a language model (LM) of the uncorrected genomic reads, and then, based on this LM, calculates a metric called theperplexity metricto evaluate the corrected reads for different parameter choices. Next, it finds the one that produces the highest alignment ratewithoutusing a reference genome. The fundamental intuition of our approach is that the perplexity metric is inversely correlated with the quality of the assembly after error correction. Therefore,Lernaleverages the perplexity metric for automated tuning ofk-mer sizes without needing a reference genome.ResultsFirst, we show that the bestk-mer value can vary for different datasets, even for the same EC tool. This motivates our design that automatesk-mer size selection without using a reference genome. Second, we show the gains of our LM using its component attention-based transformers. We show the model’s estimation of the perplexity metric before and after error correction. The lower the perplexity after correction, the better thek-mer size. We also show that the alignment rate and assembly quality computed for the corrected reads are stronglynegativelycorrelated with the perplexity, enabling the automated selection ofk-mer values for better error correction, and hence, improved assembly quality. We validate our approach on both short and long reads. Additionally, we show that our attention-based models have significant runtime improvement for the entire pipeline—18$$\times$$×faster than previous works, due to parallelizing the attention mechanism and the use of JIT compilation for GPU inferencing.ConclusionLerna improvesde novogenome assembly by optimizing EC tools. Our code is made available in a public repository at:https://github.com/icanforce/lerna-genomics.