Toghiani Sajjad, Aggrey Samuel E, Rekaya Romdhane
Animal Genomics and Improvement Laboratory, Agricultural Research Service, USDA, Beltsville, MD 20705, USA.
Institute of Bioinformatics, The University of Georgia, Athens, GA 30602, USA.
Genes (Basel). 2025 May 10;16(5):563. doi: 10.3390/genes16050563.
BACKGROUND/OBJECTIVES: Genomic selection (GS) has improved accuracy compared to traditional methods. However, accuracy tends to plateau beyond a certain marker density. Prioritizing influential SNPs could further enhance the accuracy of GS. The fixation index (F) allows for the identification of SNPs under selection pressure. Although the F method was shown to be able to prioritize SNPs across the whole genome and to increase accuracy, its performance could be further improved by focusing on the prioritization process within QTL regions.
A trait with heritability of 0.1 and 0.4 was generated under different simulation scenarios (number of QTL, size of SNP windows around QTL, and number of selected SNPs within a QTL region). In total, six simulation scenarios were analyzed. Each scenario was replicated five times. The population comprised 30K animals from the last 2 generations (G9-G10) of a 10-generation (G1-G10) selection process. All animals in G9-10 were genotyped with a 600K SNP panel. F scores were calculated for all 600K SNPs. Two prioritization scenarios were used: (1) selecting the top 1% SNPs with the highest F scores, and (2) selecting a predetermined number of SNPs within each QTL window. GS accuracy was evaluated using the correlation between true and estimated breeding values for 5000 randomly selected animals from G10.
Prioritizing SNPs using F scores within QTL window regions increased accuracy by 5 to 18%, with the 50-SNP windows showing the best performance.
The increase in GS accuracy warrants the testing of the algorithm when the number and position of QTL are unknown.
背景/目的：与传统方法相比，基因组选择（GS）提高了准确性。然而，超过一定的标记密度后，准确性往往趋于平稳。对有影响力的单核苷酸多态性（SNP）进行优先级排序可以进一步提高GS的准确性。固定指数（F）有助于识别处于选择压力下的SNP。尽管F方法已被证明能够对全基因组的SNP进行优先级排序并提高准确性，但通过关注数量性状基因座（QTL）区域内的优先级排序过程，其性能可以进一步提高。
在不同的模拟场景（QTL数量、QTL周围SNP窗口大小以及QTL区域内选择的SNP数量）下生成遗传力为0.1和0.4的性状。总共分析了六种模拟场景。每个场景重复五次。群体由来自十代（G1 - G10）选择过程中最后两代（G9 - G10）的30K只动物组成。G9 - 10中的所有动物都用600K SNP芯片进行了基因分型。计算了所有600K SNP的F分数。使用了两种优先级排序方案：（1）选择F分数最高的前1%的SNP，以及（2）在每个QTL窗口内选择预定数量的SNP。使用来自G10的5000只随机选择动物的真实育种值与估计育种值之间的相关性来评估GS准确性。
在QTL窗口区域内使用F分数对SNP进行优先级排序可将准确性提高5%至18%，其中50个SNP的窗口表现最佳。
当QTL的数量和位置未知时，GS准确性的提高保证了对该算法进行测试。
