Recombinant inbred lines derived from wide crosses in Pisum

International audienceGenomic resources are becoming available for Pisum but to link these to phenotypic diversity requires well marked populations segregating for relevant traits. Here we describe two such resources. Two recombinant inbred populations, derived from wide crosses in Pisum are described. One high resolution mapping population involves cv Caméor, for which the first pea whole genome assembly was obtained, crossed to JI0281, a basally divergent P. sativum sativum landrace from Ethiopia. The other is an inter sub-specific cross between P. s. sativum and the independently domesticated P. s. abyssinicum. The corresponding genetic maps provide information on chromosome level sequence assemblies and identify structural differences between the genomes of these two Pisum subspecies. In order to visualise chromosomal translocations that distinguish the mapping parents, we created a simplified version of Threadmapper to optimise it for interactive 3-dimensional display of multiple linkage groups. The genetic mapping of traits affecting seed coat roughness and colour, plant height, axil ring pigmentation, leaflet number and leaflet indentation enabled the definition of their corresponding genomic regions. The consequence of structural rearrangement for trait analysis is illustrated by leaf serration. These analyses pave the way for identification of the underlying genes and illustrate the utility of these publicly available resources. Segregating inbred populations derived from wide crosses in Pisum, together with the associated marker data, are made publicly available for trait dissection. Genetic analysis of these populations is informative about chromosome scale assemblies, structural diversity in the pea genome and has been useful for the fine mapping of several discrete and quantitative traits. Recombinant inbred lines (RILs) were first developed for mouse genetics 1 but are widely used in plant genetics where self-fertilization makes their development relatively straightforward. RILs capture genetic variation in a stable way. As inbred lines they are amenable to multiple investigations, such as replicated measurement or the accumulation of data over time. There are two disadvantages to RILs: they do not capture information about dominance unless this is recorded in early generations and they segregate only for the alleles that distinguish the two parents. The latter disadvantage is overcome by linkage disequilibrium mapping methods such as Multi-parent Advanced Generation Inter-Cross (MAGIC) populations 2 , Nested Association Mapping (NAM) 3 which, together with Genome-Wide Association Studies (GWAS), enable analyses of diverse populations 4 and can capture the contribution of multiple alleles. Nevertheless, sufficiently large RIL populations can provide a high degree of resolution in genetic mapping and, when the parents are sufficiently divergent, RILs can capture many bi-allelic differences. Here we present a preliminary analysis of two RIL populations derived from two wide crosses in Pisum. A wide cross within P. sativum sativum is represented by RILs derived from the cross between cv Caméor (a French field pea variety, also designated JI3253) and JI0281 (a P. s. sativum accession from Ethiopia, designated 'P. sativum landrace DCG0248' in Kreplak et al. 5). The second is an inter-specific or inter-subspecific cross. One parent of this second wide cross is JI2202 (designated 'P. sativum abyssinicum_Landrace_DCG0563 by Kreplak et al. 5) which represents a closely related group of peas that have been domesticated independently from P. s. sativum 6,7. Sometimes P. s. abyssinicum is regarded as a distinct species rather than a subspecies of P. sativum. The second parent of this population, JI2822, is a genetic stock, a RIL derived from the cross between JI0015 and JI0399, which has been widely used in mutagenesis experiments (see 8 and references therein) or for gene conten