here - First Legume Society Conference (LSC1)

Book of Abstracts First Legume Society Conference 2013: A Legume Odyssey Novi Sad, Serbia, 9-11 May 2013
_________________________________________________________________________________________
Lucerne genomics, potential application in breeding
Bernadette Julier 1 , Jérôme Gouzy², Frédéric Debellé², Philippe Barre 1
1 INRA, UR4, URP3F, Lusignan, France
2 INRA, UMR2594/441, LIPM, Castanet-Tolosan, France
Lucerne genomics is hampered by the autotetraploidy and heterozygozity of the genome.
Nevertheless, next generation sequencing (NGS) technologies now offer the opportunity to get
enough sequence data to assemble lucerne genome have haplotype information. We have used
Illumina Hiseq technology to sequence paired-end and mate-pair libraries from one lucerne
genotype. The sequences are assembled based on M. truncatula genome sequence. This reference
genome sequence will be essential for the development of high-throughput markers: SNP or
genotyping by sequencing. Lucerne genome sequence will be used to increase genetic progress in
breeding programs. A first strategy could be to build, rapidly and at low cost, dense genetic maps
for QTL identificationin multiple mapping populations. Positive alleles with strong effects could
be selected in these populations in order to increase their frequencies in the next generation. A
second strategy could be based on candidate genes known to be involved in specific trait
variation. From the sequence of these genes in lucerne, association mapping could be performed
in synthetic varietiesto identify and select positive alleles. With sufficient number of markers, a
third strategy could be to carry out genomic selection, i.e. to predict plant phenotypes from their
genotypes by using a calibration established on a training population. All these marker-assisted
selection methods combined with phenotypic selection could significantly increase genetic
progress that is relatively low in an autotetraploid species whose varieties are synthetic
populations. Indeed they offer the opportunity to select four copies of a positive allele in a single
individual instead of a single (dominant) allele with phenotypic selection.
154