Manisha Saini ¹ Raju R. Yadav ¹ Rahul Kumar ² Subhash Chandra ^3* N K. Rathod ^1* Meniari Taku ^1* Manu Yadav ¹ Sudipta Basu ^1* Ambika Rajendran ¹ S K Lal ¹ Akshay Talukdar ^1*

• ¹ Indian Agricultural Research Institute (ICAR), New Delhi, India
• ² ICAR-Resarch complex for NEH Region Tripura centre lembucherra, Tripura, India
• ³ Central institute of cotton resesarch, RS Sirsa, Sirsa, Haryana, India

Soybean [Glycine max (L.) Merrill] contains 18-20% oil, 40-45% protein and countless nutrients vital for human health. It is grown worldwide for food, feed, pharmaceutical and industrial applications. However, inherent loss of seed viability during ambient storage poses serious bottleneck in the production and maintenance of quality seeds.In this study, an attempt was made to map QTLs and identify candidate genes for seed viability in soybean. A high viable genotype EC1023 (>90% germination after one year of storage) was hybridized with VLS61, a poor storing genotype (<70% germination after one year of storage), and the F1 seeds were advanced to the next generation. The F2:3 seeds were subjected to accelerated ageing (AA) followed by viability testing through germination test. The germination of the parental genotypes EC1023 and VLS61 were 40% and 14%, respectively, and that of the F2:3 seeds ranged from 4.16% to 71.42 % indicating wide variability in the viability of the seeds. Genetic polymorphism studied with 506 SSR markers indicated the polymorphism between the parental genotypes to be 20.35%; however, distribution of the polymorphism was not uniform across the chromosomes; Chr. 14 had 30.00 % polymorphism as against 7.14 % on Chrs. 12. Through QTL ICIM approach, 8 QTL for seed viability were mapped on viz., 6, 7, 8, 10, 13 and 17, respectively. The phenotypic variation of the QTLs were 1.97-11.10%. Two QTL viz., qSv-7.1 (PVE=11.10%) and qSv13.1 (PVE=11.08%) appeared to be major QTLs for seed viability and rest minor ones. All QTL except qSv8.2 appeared to be novel. The mapped QTLs were validated in 40 interspecific RILs with varying level of seed viability. The SSR markers Satt538 linked to the QTL qSv8.2 could successfully (70%) separate the highly viable RILs from the poor-viable RILs. Similarly, SSR markers Sat_316 and Sat_173 were successful by 80-85% in separating the high and poor viable RILs. Based on PANTHER and gene annotation information,more than 500 candidate genes for seed viability underlying the mapped QTL were identified. The mapped QTL and the identified candidate genes will pave the way for marker-assisted breeding of soybean to generate genotypes with improved seed viability.