Hebei Key Laboratory of Forensic Medicine, College of Forensic Medicine, Hebei Medical University, Shijiazhuang 050017, China.
Fa Yi Xue Za Zhi. 2023 Jun 25;39(3):262-270. doi: 10.12116/j.issn.1004-5619.2023.530107.
To compare the application value of the likelihood ratio (LR) method and identity by state (IBS) method in the identification involving half sibling relationships, and to provide a reference for the setting of relevant standards for identification of half sibling relationship.
(1) Based on the same genetic marker combinations, the reliability of computer simulation method was verified by comparing the distributions of cumulated identity by state score (CIBS) and combined full sibling index in actual cases with the distributions in simulated cases. (2) In different numbers of three genetic marker combinations, the simulation of full sibling, half sibling and unrelated individual pairs, each 1 million pairs, was obtained; the CIBS, as well as the corresponding types of cumulative LR parameters, were calculated. (3) The application value of LR method was compared with that of IBS method, by comparing the best system efficiency provided by LR method and IBS method when genetic markers in different amounts and of different types and accuracy were applied to distinguish the above three relational individual pairs. (4) According to the existing simulation data, the minimum number of genetic markers required to distinguish half siblings from the other two relationships using different types of genetic markers was estimated by curve fitting.
(1) After the rank sum test, under the premise that the real relationship and the genetic marker combination tested were the same, there was no significant difference between the simulation method and the results obtained in the actual case. (2) In most cases, under the same conditions, the system effectiveness obtained by LR method was greater than that by IBS method. (3) According to the existing data, the number of genetic markers required for full-half siblings and half sibling identification could be obtained by curve fitting when the system effectiveness reached 0.95 or 0.99.
When distinguishing half sibling from full sibling pairs or unrelated pairs, it is recommended to give preference to the LR method, and estimate the required number of markers according to the identification types and the population data, to ensure the identification effect.
比较似然比（LR）法和状态一致法（IBS）在半同胞关系鉴定中的应用价值，为半同胞关系鉴定相关标准的制定提供参考。
（1）基于相同的遗传标记组合，通过比较实际案例与模拟案例中累积状态一致评分（CIBS）和组合全同胞指数的分布，验证计算机模拟方法的可靠性。（2）在不同数量的 3 个遗传标记组合中，分别模拟全同胞、半同胞和无关个体对，每个模拟 100 万对；计算 CIBS 以及相应类型的累积 LR 参数。（3）通过比较 LR 法和 IBS 法在不同数量和类型及精度的遗传标记应用于区分上述 3 种关系个体对时所能提供的最佳系统效率，比较 LR 法的应用价值。（4）根据现有的模拟数据，通过曲线拟合估计使用不同类型遗传标记区分半同胞与其他两种关系所需的最小遗传标记数量。
（1）秩和检验后，在真实关系和测试的遗传标记组合相同的前提下，模拟方法与实际案例结果无显著差异。（2）在大多数情况下，在相同条件下，LR 法获得的系统效率大于 IBS 法。（3）根据现有数据，当系统效率达到 0.95 或 0.99 时，可以通过曲线拟合获得全半同胞和半同胞识别所需的遗传标记数量。
在区分半同胞与全同胞个体对或无关个体对时，建议优先选择 LR 法，并根据鉴定类型和人群数据估计所需的标记数量，以保证鉴定效果。
