RAfilter: an algorithm for detecting and filtering false-positive alignments in repetitive genomic regions

Abstract

Telomere to telomere (T2T) assembly relies on the correctness of sequence alignments. However, the existing aligners tend to generate a high proportion of false-positive alignments in repetitive genomic regions which impedes the generation of T2T-level reference genomes for more important species. In this paper, we present an automatic algorithm called RAfilter for removing the false-positives in the outputs of existing aligners. RAfilter takes advantage of rare k-mers representing the copy-specific features to differentiate false-positive alignments from the correct ones. Considering the huge numbers of rare k-mers in large eukaryotic genomes, a series of high-performance computing techniques such as multi-threading and bit operation are used to improve the time and space efficiencies. The experimental results on tandem repeats and interspersed repeats show that RAfilter was able to filter 60%–90% false-positive HiFi alignments with almost no correct ones removed, while the sensitivities and precisions on ONT datasets were about 80% and 50% respectively.

Acknowledgements

This work was supported by Shenzhen Science and Technology Program (Grant No. RCBS20210609103819020) and National Natural Science Foundation of China (Grant No. 32100501). The authors thank Shengcheng Zhang for the discussions about high-efficient implementation.

Author Contributions

W.P., J.Y. and Y.Y. designed this study. J.Y., X.Z. and H.J developed and tested RAfilter. J.Y., Y.Y and W.P. wrote the paper.

Data Availability

The software, source codes and manual of RAfilter can be freely downloaded at https://github.com/panlab-bioinfo/RAfilter.git.

Conflict of interests statement

None declared.

References

[1] Sohn J, Nam JW . The present and future of de novo whole genome assembly. Brief Bioinform. 2018; 19: 23-40.

[2] Hon T, Mars K, Young G et al. Highly accurate long-read HiFi sequencing data for five complex genomes. Sci Data. 2020; 7: 399.

[3] Cheng H, Concepcion GT, Feng X et al. Haplotype-resolved de novo assembly using phased assembly graphs with hifiasm. Nat Methods. 2021; 18: 170-5.

[4] Kolmogorov M, Yuan J, Lin Y et al. Assembly of long, error-prone reads using repeat graphs. Nat Biotechnol. 2019; 37: 540-6.

[5] Nurk S, Walenz BP, Rhie A et al. HiCanu: accurate assembly of segmental duplications, satellites, and allelic variants from high-fidelity long reads. Genome Res. 2020; 30: 1291-305.

[6] Ruan J, Li H . Fast and accurate long-read assembly with wtdbg2. Nat Methods. 2020; 17: 155-8.

[7] Nurk S, Koren S, Rhie A et al. The complete sequence of a human genome. Science. 2022; 376: 44-53.

[8] Deng Y, Liu S, Zhang Y et al. A telomere-to-telomere gap-free reference genome of watermelon and its mutation library provide important resources for gene discovery and breeding. Mol Plant. 2022; 15: 1268-84.

[9] Mascher M, Wicker T, Jenkins J et al. Long-read sequence assembly: a technical evaluation in barley. Plant Cell. 2021; 33: 1888-906.

[10] Hou X, Wang D, Cheng Z et al. A near-complete assembly of an Arabidopsis thaliana genome. Mol Plant. 2022; 15: 1247-50.

[11] Li H. New strategies to improve minimap2 alignment accuracy. Bioinformatics. 2021; 37: 4572-4.

[12] Marcais G, Kingsford C . Jellyfish: a fast k-mer counter. Tutorialis E Manuais. 2012; 1: 1-8.

[13] Ono Y, Asai K, Hamada M . PBSIM: PacBio reads simulator-toward accurate genome assembly. Bioinformatics. 2013; 29: 119-21.

[14] Ono Y, Asai K, Hamada M . PBSIM2: a simulator for long-read sequencers with a novel generative model of quality scores. Bioinformatics. 2021; 37: 589-95.

[15] Zhang Y, Fu J, Wang K et al. The telomere-to-telomere gap-free genome of four rice parents reveals SV and PAV patterns in hybrid rice breeding. Plant Biotechnol J. 2022; 20: 1642-4.