Comparison of the likelihood ratio and identity-by-state scoring methods for analyzing sib-pair test cases: a study...

Area of Science:

• Forensic Genetics
• Population Genetics

Background:

• Sibling relationship testing is crucial in forensic and genetic studies.
• Likelihood ratio (LR) and identity-by-state (IBS) are statistical methods used for relationship inference.
• Assessing the discriminatory power of DNA markers is essential for accurate genetic testing.

Purpose of the Study:

• To evaluate the effectiveness of LR and IBS methods in sibling pair analysis.
• To compare the statistical power of LR and IBS using multiple DNA markers.
• To determine the significance of these methods in distinguishing full-siblings, half-siblings, and unrelated individuals.

Main Methods:

• Computer simulations were conducted using 10 DNA markers (HLA-DQalpha, D1S80, 8 STR loci).
• Generated pedigrees of 10,000 pairs for full-siblings, half-siblings, and unrelated individuals.
• Analyzed data using both summation of log10 LR and IBS scoring for each group.

Main Results:

• Both LR and IBS methods showed significant differences between sibling groups.
• LR method achieved 91% power to detect unrelated individuals and 38% for half-siblings (vs. full-siblings) at a 5% false-positive rate.
• IBS method showed 87% and 42% powers, respectively; LR exceeded 100 in 62% of full-siblings.

Conclusions:

• The LR method offers a convincing advantage over IBS for identifying unrelated individuals as non-siblings.
• Both LR and IBS methods are valuable for sibling pair testing, especially with numerous polymorphic markers.
• LR method provides a high probability of full-sibship confirmation, particularly when LR values exceed 100.