P0769
Construction of High Quality Recombination Maps in Sorghum Bicolor using Recombinant Inbred Line Populations

Santosh Deshpande , ICRISAT, HYDERABAD, India
Sivasubramani Selvanayagam , ICRISAT, Hyderababd, India
Ramu Punna , Institute for Genomic Diversity, Cornell University, Ithaca, NY
CT Hash , ICRISAT, Niamey, Niger
Hari D Upadhyaya , ICRISAT, Hyderabad, India
Geoffrey Morris , Kansas State University, Manhattan, KS
Charlotte Acharya , Cornell University, Ithaca, NY
Sharon E. Mitchell , Cornell University, Ithaca, NY
Edward S. Buckler , USDA-ARS-Cornell University, Ithaca, NY
Stephen Kresovich , Clemson University, Clemson, SC
Genotyping-By-Sequencing (GBS) is a cost effective high-throughput genotyping approach, produces large number of Single Nucleotide Polymorphisms (SNPs) on low coverage genomic data, which is being widely used in crop improvement. At ICRISAT-HQ, Patancheru, GBS experiment was carried out for Sorghum biocolor (L.) Moench Recombinant Inbred Line (RIL) populations to construct high quality recombination maps with help of available Sorghum bicolor (BTx623) reference annotations. A total of 2808 progenies from six RIL populations genotyped with GBS approach were considered for this project. The RIL populations segregated for several biotic stresses (grain mold, shoot fly, striga and stem borer) and post-flowering drought tolerance. Mapped data was filtered for taxa coverage (progenies with <10 percent) and site coverage (with <25 percent) were eliminated from the analysis, and final data with 2561 individuals with circa 100k SNPs was generated. We imputed/corrected the missing SNP calls using FSFHap (TASSEL5.0 plugin) and total polymorphic SNP calls across RIL sets ranged from 15k to 22k. We propose to construct recombination maps from polymorphic SNPs for individual RIL population by combining parental calls into individual SNP blocks following the parental pedigree. Recombination breakpoints will be defined to construct skeleton bins. To eliminate false double recombinants, we are testing sliding block window approach with differential sets of 1000 to 2000 bases containing minimal number of SNPs. The constructed recombination maps could provide a useful basis for analyzing genome-wide variation of complex traits.