Sampling bias in estimating Design II variance components with S1 families

Area of Science:

• Plant breeding
• Quantitative genetics
• Statistical genetics

Background:

• Design II mating schemes are useful for estimating genetic variances in plants.
• Plants with single pistillate flowers necessitate using S1 individuals to represent S0 individuals in mating designs.
• Previous methods did not fully account for bias introduced by using S1 lines.

Purpose of the Study:

• To investigate the bias in additive (V A) and dominance (V D) variance estimates when using S1 lines in Design II mating.
• To develop a method for correcting these biased estimates.
• To highlight the impact of additive genetic variance on dominance variance estimates.

Main Methods:

• Utilized a Design II mating scheme with S1 individuals representing S0 individuals.
• Employed a small number of individuals (e.g., 10) per S1 line.
• Derived formulas to compute and correct for upward bias in V A and V D estimates.
• Analyzed the influence of the number of individuals used (m1 and m2) on bias.

Main Results:

• Estimates of V A and V D are biased upwards when using a small number of S1 individuals per line.
• Additive genetic variance (V A) was found to contribute to the bias in dominance variance (V D) estimates.
• Correction factors can be computed to adjust the biased estimates.
• Bias is dependent on the number of individuals used to represent S1 lines in half-sib and full-sib families (m1 and m2).

Conclusions:

• The proposed method allows for accurate estimation of genetic variances even with the use of S1 lines in Design II.
• Corrected estimates of V A and V D are crucial for effective plant breeding programs.
• Understanding and correcting for bias ensures more reliable genetic parameter estimates.