Data Standards for Multiplex Amplicon Sequencing

Data Standards for Multiplex Amplicon Sequencing Chris Kent 2 Overview •Background •Amplicon Sequencing •PrimalScheme •The problem •How we caused it •A solution •How we attempted to solve it 3 Research question? • How do I recover “this” pathogen’s genome from a complex sample? •Incredibly low “target” concentration •High levels of background •How do I do this cheaply, and as accessibly as possible? •LMICs carry the burden of infectious disease 4 What is amplicon sequencing? •Multiplexed PCR •Uses PCR to amplify the target nucleotides •Very specific •Can work down to the limit of detection (1-10 copies per µL) •Very cheap •Incredibly sensitive, cheap and accessible. 5 What is amplicon sequencing? We use overlapping amplicons to amplify sequences of interest •2 or more pools are needed to prevent overlap between products 6 What data standards does PCR need? •Primer Sequences •Required for ordering producing the oligonucleotides 7 What data standards does PCR need? •Primer Sequences •Required for ordering producing the oligonucleotides •Organisation of the primers •Which PCR pool does each go into? 8 What data standards does PCR need? •Primer Sequences •Required for ordering producing the oligonucleotides •Organisation of the primers •Which PCR pool does each go into? •Indexing system •Primers are synthetic DNA and need to be removed from read data 9 History •The Quick Lab uses tiling amplicon sequencing (and brand-new sequencing tech) to sequence EBV in field 2016 Quick, J. et al. (2016) ‘Real-time, portable genome sequencing for Ebola surveillance’, Nature, 530(7589), pp. 228–232. Available at: https://doi.org/10.1038/nature16996. 16 Changes to the file • First scheme “nCoV-19 V1” •Sequence is missing? •Bioinformatic processing to get sequences https://github.com/artic-network/primerschemes/blob/master/nCoV-2019/V1/nCoV-2019.primer.bed 17 Changes to the file • First scheme ”nCoV-19 V1” •Sequence is missing? •Bioinformatic processing to get sequences •Any differences to the reference genome https://github.com/artic-network/primerschemes/blob/master/nCoV-2019/V1/nCoV-2019.primer.bed 18 Changes to the file • First scheme ”nCoV-19 V1” •Sequence is missing? •Bioinformatic processing to get sequences •Any differences to the reference genome https://github.com/artic-network/primerschemes/blob/master/nCoV-2019/V4/nCoV-2019.primer.bed 19 Changes to the file • First scheme ”nCoV-19 V1” •Sequence is missing? •Bioinformatic processing to get sequences •Any differences to the reference genome •Multiple primers •SARS-CoV-2_10_LEFT_alt1 https://github.com/artic-network/primerschemes/blob/master/nCoV-2019/V4.1/nCoV-2019.primer.bed 20 Changes to the file •We needed to encode additional data •To improve performance schemes needed rebalancing Flexibility for the future •Ensure future use cases can be accounted for https://github.com/artic-network/primerschemes/blob/master/nCoV-2019/V4.1/nCoV-2019.primer.bed 21 Metadata changes •Lack of versioning •v1￫ v2￫ v3￫ v4￫ v4.1￫ v5.3.2￫ v5.4.2 •Lack of citable name •ARTIC v4.1 •v4.1 •Covid v4.1 •nCoV v4.1 No encompassing metadata standard 22 First step to fixing •Formalised the primer scheme specification •Developed tooling for parsing •PrimalBedTools doi:10.5281/zenodo.16366659 23 First step to fixing •Formalised the spec •Primer Level key:value •“{k1}={v1};{k2}={v2}” •Primer Class •“LEFT | RIGHT | PROBE” •Structured PrimerName •{scheme}_{amplicon#}_{class}_{suffix} doi:10.5281/zenodo.16366659 24 First step to fixing •Formalised the spec •Primer Level key:value •“{k1}={v1};{k2}={v2}” •Primer Class •“LEFT | RIGHT | PROBE” •Structured PrimerName •{scheme}_{amplicon#}_{class}_{suffix} doi:10.5281/zenodo.16366659 25 Metadata standard •Primerschemes open access repository •{schemename}/{ampliconsize}/{version} •Versioning v{x}.{y}.{z} •X – New Scheme •Y – Change to the primers •Z – Change to attributes https://github.com/quick-lab/primerschemes 32 Future work •This has been a long running collaboration with PHA4GE •File specification is a superset of old files •Primal-page (labs) is not •Slightly opinionated •Hosted by ARTICnetwork On going work to consolidate both + formalise primaschema primal-page 33 Acknowledgements ARTICnetwork + quick lab Scripts Research •Bede Constantinides * •Andy Smith •Josh Quick B&M Gates Foundation PHA4GE BCCDC