Edit me

Step 2.2 Artifact removal: local realignment around indels

Some artifacts may arise due to the alignment stage, especially around indels where reads covering the start or the end of an indel are often incorrectly mapped. This results in mismatches between the reference and reads near the misalignment region, which can easily be mistaken for SNPs. Thus, the realignment stage aims to correct these artifacts by transforming those regions with misalignment due to indels into reads with a consensus indel for correct variant calling.

Realignment can be accomplished using the GATK IndelRealigner 1 (https://software.broadinstitute.org/gatk/documentation/tooldocs/3.8-0/org_broadinstitute_gatk_tools_walkers_indels_IndelRealigner.php ). Alternatives include Dindel 2 (https://github.com/genome/dindel-tgi ) and SRMA 3 (https://github.com/nh13/SRMA ).

The inclusion of the realignment stage in a variant calling pipeline depends on the variant caller used downstream. This stage might be of value when using non-haplotype-aware variant caller like the GATK’s UnifiedGenotyper 4. However, if the tool used for variant calling is haplotype-aware like Platypus 5, FreeBayes 6 or the HaplotypeCaller 7, then it is not needed nor recommended. The GATK recommendations starting from their 3.6 release onwards, and the guidelines for functional equivalence 8 also vote against this stage.

Ultimately however, characteristics of the dataset at hand would dictate whether realignment and other clean-up stages are needed. Ebbert et al paper for example argues against PCR duplicates removal 9, while Olson et al recommends all the stages of clean up applied to the dataset at hand 10. Some experimentation is therefore recommended when handling real datasets.

Bibliography

  1. DePristo, M. A. et al. A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nat. Genet. 43, 491–498 (2011). ↩

  2. Albers, C. A. et al. Dindel: accurate indel calls from short-read data. Genome Res. 21, 961–973 (2011). ↩

  3. Homer, N. & Nelson, S. F. Improved variant discovery through local re-alignment of short-read next-generation sequencing data using SRMA. Genome Biol. 11, R99 (2010). ↩

  4. Van der Auwera, G. A. et al. From FastQ data to high confidence variant calls: the Genome Analysis Toolkit best practices pipeline. Curr. Protoc. Bioinformatics 11, 11.10.1-11.10.33 (2013). ↩

  5. Rimmer, A. et al. Integrating mapping-, assembly- and haplotype-based approaches for calling variants in clinical sequencing applications. Nat. Genet. 46, 912–918 (2014). ↩

  6. Garrison, E. & Marth, G. Haplotype-based variant detection from short-read sequencing. arXiv (2012). ↩

  7. Poplin, R. et al. Scaling accurate genetic variant discovery to tens of thousands of samples. BioRxiv (2017). ↩

  8. Regier, A. A. et al. Functional equivalence of genome sequencing analysis pipelines enables harmonized variant calling across human genetics projects. BioRxiv (2018). doi:10.1101/269316 ↩

  9. Ebbert, M. T. W. et al. Evaluating the necessity of PCR duplicate removal from next-generation sequencing data and a comparison of approaches. BMC Bioinformatics 17 Suppl 7, 239 (2016). ↩

  10. Olson, N. D. et al. Best practices for evaluating single nucleotide variant calling methods for microbial genomics. Front. Genet. 6, 235 (2015). ↩





Tags: genomics_analysis