A comparative evaluation of hybrid error correction methods for error-prone long reads

Abstract

BackgroundThird-generation sequencing technologies have advanced the progress of the biological research by generating reads that are substantially longer than second-generation sequencing technologies. However, their notorious high error rate impedes straightforward data analysis and limits their application. A handful of error correction methods for these error-prone long reads have been developed to date. The output data quality is very important for downstream analysis, whereas computing resources could limit the utility of some computing-intense tools. There is a lack of standardized assessments for these long-read error-correction methods.ResultsHere, we present a comparative performance assessment of ten state-of-the-art error-correction methods for long reads. We established a common set of benchmarks for performance assessment, including sensitivity, accuracy, output rate, alignment rate, output read length, run time, and memory usage, as well as the effects of error correction on two downstream applications of long reads: de novo assembly and resolving haplotype sequences.ConclusionsTaking into account all of these metrics, we provide a suggestive guideline for method choice based on available data size, computing resources, and individual research goals.Electronic supplementary materialThe online version of this article (10.1186/s13059-018-1605-z) contains supplementary material, which is available to authorized users.

Obstacles to Studying Alternative Splicing Using scRNA-seq

“…Long read technologies could in theory enable 100% accurate isoform quan-tification, if issues due to a high base calling error rate could be overcome [23]. However we find that even when no isoform detection errors occur, our ability to accurately detect isoforms is very limited.…”

Chromonomer: a tool set for repairing and enhancing assembled genomes through integration of genetic maps and conserved synteny

“…Long-read sequencing technologies, particularly with the introduction of the Sequel I and II platforms (Pacific Biosciences), now have high enough volume that new assemblies are being constructed purely from long reads. Although long reads currently have high indel-derived error rates, those errors can be corrected with sufficient sequencing depth (Fu et al 2019). These assemblies have huge N50 measures, and typically contain very few or even no gaps because very limited or no scaffolding is employed, although repeats occasionally result in mis-joins of contigs.…”

A New Way to Discover IRESs in Pathology or Stress Conditions? Harnessing Latest High‐Throughput Technologies

“…Second, the random error rate of TGS was reported to be more than 10%, which is higher than that in NGS. [44] This may affect the results of downstream analyses, including the accuracy of genetic variation identification. To resolve this problem, the accuracy could be improved to more than 99% by increasing the sequencing depth by using more SMRT cells.…”