單親親子鑑定中LR與RMNE分布之比較研究

LR and RMNE Distributions in Single-Parentage Testing

10.6134/tjfm.2009.01.01.10

李俊億(James Chun-I Lee)

親子鑑定 ； 似然比 ； 不能排除的隨機男子 ； 短相連重複DNA基因 ； Parentage testing ； LR ； RMNE ； STR loci

台灣法醫學誌

1卷1期（2009 / 06 / 01）

41 - 48

繁體中文

本研究針對親子二人組探討其LR與RMNE演算模式間的差異現象。從具有15個STR基因(CSF1PO、D2S1338、D3S1358、D5S818、D7S820、D8S1179、D13S317、D16S539、D18S511、D19S433、D21S11、FGA、TH01、TPOX、vWA)的一萬個模擬家庭中，隨機抽樣具有親子關係的1244個二人組親子組合，分別進行15個STR基因之親子關係LR與RMNE演算。本研究將RMNE值轉爲1/RMNE，成爲系爭父親與不能排除的隨機男子機率之比值，以利比較分析。依據統計結果，模擬的1244個親子樣品其LR與1/RMNE數據均呈現常態分配。結果顯示LR數據的分散程度明顯大於1/RMNE的分布，LR的平均數大於1/RMNE的平均數6倍以上。在個案的LR與1/RMNE數據的比值差異上，差異從0.0158倍到34191.50倍，平均是8.56倍。此數據顯示，由於演算方法不同，可能產生的結果差異極大，因此必須謹慎。由於LR法將參與鑑定的各方基因型都列入計算而受提供必備對偶基因型的機率與頻率影響，因此，在同一案例中以LR與1/RMNE表示的數據差異極大。在1/RMNE數據的變化上，由於影響無法排除的頻率關鍵在無法排除的對偶基因型數目而非頻率，因此，1/RMNE數據的變化範圍較小，將隨著篩檢基因數目增加，而有規律的減少無法排除的頻率。

This study discovered the distributed features of LR and RMNE values of parentage testing. From 10,000 simulated families with 15 STR loci (CSF1PO、D2S1338、D3S1358、D5S818、D7S820、D8S1179、D13S317、D16S539、D18S511、D19S433、D21S11、FGA、TH01、TPOX、vWA), a total number of 1244 duo families were randomly picked for LR and RMNE calculations. The value of RMNE was transferred to 1/RMNE which stands for the likelihood ratio between the probabilities of alleged father and random man not excluded. From the results of statistics, distributions of LR and 1/RMNE data from duo families are normal curves. The distribution of LR data is more dispersed than that of 1/RMNE. And the mean of LR data is 6 times larger than that of 1/RMNE. From the ratio of LR and 1/RMNE of individual cases, the data ranged from 0.0158 to 34191.50. These data show that the different mathematical methods will seriously affect the results. Since the LR was calculated using genotypes of all join members of the test which was affected by the contributed probability and frequencies of obligated alleles, it caused a large variation between LR and 1/RMNE expressions in the same case. For the 1/RMNE, since the number, not the frequency, of obligated alleles of the biological father is the main factor to affect the data, it will not cause much variation. The value of RMNE is stable and gradually decreased when more STR loci are screened.

1. SigmaPlot 2004 for Windows version 9.01, Systat Software, Inc..
2. Brenner, CH(1993).A note on paternity computation in cases lacking a mother.Transfusion,33,51-4.
3. Siegel, JA(ed.),Knupfer, GC(ed.),Saukko, PJ(ed.)(2000).Encyclopedia of Forensic Sciences.Academic Press.
4. Walker, RH(ed.)Duquesnoy, RJ(ed.),Jennings, ER(ed.)(1983).Inclusion probabilities in parentage testing.Arlington,Va:Am Assoc Blood. Banks.
5. 李俊億(2008)。法科學DNA比對系統。台北:全傑電腦。
6. 李俊億、張奕元、蘇志文。台灣地區漢人STR與Y-STR基因頻率數據分析。中華民國鑑識科學學會2006年鑑識科學研討會論文集
7. 李俊億、謝幸媚(2008)。親子鑑定的演算邏輯。台北:國立台灣大學出版中心。