TGS-GapCloser: fast and accurately passing through the Bermuda in large genome using error-prone third-generation long reads

Abstract

The completeness and accuracy of genome assemblies determine the quality of subsequent bioinformatics analysis. Despite benefiting from the medium/long-range information of third-generation sequencing techniques, current gap-closing tools to enhance assemblies suffer multi-alignments and high error rates, resulting in huge time and money costs. We developed a software tool, TGS-GapCloser that uses the low depth (>=10X) single molecule sequencing long reads without any error correction to close gaps. The algorithm distinguishes gap regions from the alignments of long reads against original scaffolds, corrects only the candidate regions, and assigns the best sequences to each gap. We demonstrate that TGS-GapCloser improves the contig N50 value of draft assembly by 25-fold on average, updating above 90% gaps with 93.96% positive predictive value. Despite of high error rate of raw long reads, improved assemblies archive Q50 (99.999%) single-base accuracy with only 11.8% decrement to inputs. Besides it could complete more gaps, and is also ~29-fold faster than mainstream gapclosing tools. BUSCO analysis revealed that 3.4%-13.1% more expected genes were complete. TGS-GapCloser also shows its power to fill gaps for ultra large genome assembly of ginkgo (~12Gb) with 71.6% of gaps closed. The validation of inserted or merged gap sequences was conducted with NGS reads and reference genomes, respectively. The updated genome assemblies may promote the gene annotation, structure variant calling and thus improving the downstream analysis of ontogeny, phylogeny, and evolution.

De novo sequencing and chromosomal‐scale genome assembly of leopard coral grouper, Plectropomus leopardus

“…To further improve the integrity and accuracy of the genome assembly, we employed tgs‐gapcloser , which uses low‐depth (≥10×) single‐molecule sequencing long reads without any error correction to close gaps in the draft assembly (Xu et al, 2019). The long sequences were split into three groups, including total reads (with options –min_idy 0.2, –min_match 200 –r_round 1), reads with length ≥20 kb (with options –min_idy 0 –min_match 0 –r_round 3) and reads with length 2–20 kb (with options –min_idy 0 –min_match 0 –r_round 3), and each group was used to fill the corresponding aligned gaps.…”

Three genomes of Osteoglossidae shed light on ancient teleost evolution

“…We assembled the genome into 4,244 scaffolds which span ∼669.73 Mb (99.45% of the estimated genome size 673.41Mb) with an ultra-long scaffold N50 of ∼9.62 Mb. To improve the continuity of this assembly, we sequenced more 10.17 Gb (∼13.14-fold) single molecular long reads using Nanopore sequencing platform, resulting in a notable increasing of contig N50 value from 255.61 Kb to 2.31 Mb using TGS-GapCloser[20] ( Supplemental table 2 ). Based on the high quality draft genome assembly, we also sequenced 21.23 Gb data of a Hi-C library and anchored 647.59 Mb (∼96.80% of the whole assembly) scaffold sequences onto 20 chromosomes ( Fig.…”

“…In this process, the stLFR reads were first pre-processed to be compatibly handled by supernova assembler, using the stLFR2Supernova pipeline (https://github.com/BGI-Qingdao/stlfr2supernova_pipeline). Then, we enhanced the draft assembly using TGS-GapCloser pipeline[20] based on the single molecular long reads.…”

African Arowana Genome Provides Insights on Ancient Teleost Evolution

“…To further improve the continuity, we sequenced ∼10.2 Gb (∼13.1×) Nanopore long reads to fill the gaps. With these long reads, the contig N50 was further improved from 255.6 Kb to 2.31 Mb ( Table S2 ) using TGS-GapCloser ( Xu et al., 2019 ). To anchor the scaffold sequences to chromosomes, we constructed a Hi-C library and sequenced ∼21.2 Gb Hi-C data and thus ∼650.44 Mb sequences were anchored to 20 chromosomes ( Figures 1 A, S2 , and Table S3 ), which was consistent with the previous report on African arowana karyotype ( Oliveira et al., 2019 ).…”