Most investigations for estimating gene effects contributing to metric traits has been exclusively focused on means of earlier filial (¯F_1 and ¯F_2) and backcross populations (¯BC_11 and ¯BC_12). In this paper a novel approach is suggested using data generated from advanced filial (¯F_(3 ),¯F_(4 )….¯F_(n )) and backcross populations)¯( BC)_n1 and ¯( BC)_n2). The following set of equations is derived by algebraic solution of ¯P_1,¯P_2,¯F_3,¯F_4,¯BC_31 and ¯BC_32 means:
Mean (M) = 1/98 (4 9 ( P) ̅_1+49 P ̅_2 - 64 (BC) ̅_31-64 (BC) ̅_32+ 128 F ̅_4 ),
Additive effect (D) =□(□(□(1/(2 ))) ) (P ̅_1- P ̅_2 ),
Dominance effect (H) =1/49 (384 (BC) ̅_31+384 (BC) ̅_32 -196 F ̅_3-294 ( P) ̅_1-294 P ̅_2+ 16F ̅_4 ),
Three types of epistasis
Additive x additive effect (I) = 32/49 ( (BC) ̅_31+ (BC) ̅_32-2 F ̅_4 ),
Additive x dominance effect (J) = 1/7 (32 (BC) ̅_31-32 (BC) ̅_32-28 ( P) ̅_1+28 P ̅_2 ) and
Dominance x dominance effect (L) =16/49 (98 F ̅_3-68F ̅_4+49 ( P) ̅_1+49 P ̅_2-64 (BC) ̅_31-64 (BC) ̅_32 )
The proposed equations have been proved mathematically via theoretical working example.