Genotyping‐by‐sequencing for biogeography

Abstract

Aim: Genotyping-by-sequencing (GBS) and similar reduced-representation sequencing methods, such as restriction site-associated DNA sequencing (RADseq), have been revolutionary for genetic analyses in biogeography. However, navigating the many different methodological and analytical approaches and numerous sources of potential error can be overwhelming. We provide an overview of key considerations for biogeographical research using GBS, from sample design through data filtering to the sharing of data, which should particularly assist new users.Taxon: All taxa. Location: Worldwide. Methods:We review recent advances for GBS and compare differences among GBS methods and analytical approaches. We highlight the concerns most relevant for biogeographical research, and emphasise practical limitations for studies on non-model organisms.Results: GBS methods vary substantially and recent literature demonstrates the need for careful study design and data filtering relevant to the study organism and hypothesis under investigation. Biogeographical research using non-model organisms or long-term sampling are likely to face some practical limitations compared to ideal GBS study designs. The methodological information recorded in published manuscripts often varies. We outline a general framework for planning and undertaking biogeographical research using GBS.Main conclusions: GBS and similar approaches have grown rapidly in popularity for biogeographical research. Evaluating, recording and justifying decisions throughout a GBS workflow-across sampling, library preparation and sequencing, identifying and filtering samples and loci, biogeographical analyses, and sharing data-is crucial for improving scientific reproducibility and compatibility among GBS datasets. This review outlines ways to improve and simplify GBS research, thereby enhancing our capacity to use genomic data to address broad-scale biogeographical questions.

Low effective population size but high heterozygosity revealed by SNP analyses in adult and juvenileThuja plicatain UK Woodlands

“…Genotyping by sequencing can discover thousands of genomic variants in non-model species but requires a bioinformatic procedures for data cleaning due to genotyping errors (Mastretta-Yanes et al 2015) and missing data (López de Heredia et al 2020; Vaux et al 2023). This typically includes steps to filter or test for missing data, minor allele frequency, minimum and maximum depth, hardy-weinberg and linkage disequilibrium filtering (Vaux et al 2023), and may involve imputation of missing alleles using statistical methods (López de Heredia et al 2020; Mora-Márquez et al 2023).…”

“…Genotyping by sequencing can discover thousands of genomic variants in non-model species but requires a bioinformatic procedures for data cleaning due to genotyping errors (Mastretta-Yanes et al 2015) and missing data (López de Heredia et al 2020; Vaux et al 2023). This typically includes steps to filter or test for missing data, minor allele frequency, minimum and maximum depth, hardy-weinberg and linkage disequilibrium filtering (Vaux et al 2023), and may involve imputation of missing alleles using statistical methods (López de Heredia et al 2020; Mora-Márquez et al 2023). We found 199,704 variants before filtering, which is relatively low compared to the 1.3M SNPs found in Chinese fir ( Cunninghamia lanceolata ) also a member of the Cupressaceae family (Zheng et al 2019) or the 349,542 SNPs found in the foxtail pine ( Pinus balfouriana ) (Friedline et al 2015).…”

“…Nowadays, GBS is used to describe methods for sequencing and genotyping which use one or several restriction enzyme cutting sites (Andrews et al 2016). GBS is useful allows the direct discovery of thousands of single nucleotide polymorphisms (SNP) without prior information; however, GBS requires deep data curation and complex bioinformatic pipelines (Vaux et al 2023, Guillardín-Calvo et al 2019). For example, estimating genotyping errors (Mastretta-Yanes et al 2015) and missing data handling by imputation (Mora-Márquez et al 2023) are key steps to process GBS data to obtain reliable results.…”