Relationship inference from the genetic data on parents or offspring: A comparative study

Area of Science:

• Population genetics
• Anthropology
• Forensic science

Background:

• Relationship inference from genetic data is vital across disciplines.
• Direct inference uses couple's data; indirect inference uses offspring's data.
• The optimal sampling strategy (couples vs. offspring) for relationship studies remains debated due to cost and information trade-offs.

Purpose of the Study:

• To compare the informativeness of sampling couples versus offspring for relationship inference.
• To evaluate the performance of single-point and multi-point methods in relationship inference using different sampling strategies.
• To determine the most advantageous sampling scheme for various degrees of genetic relatedness.

Main Methods:

• Simulated 1000 trios across 10 relationship types using real human haplotypes.
• Compared genome sharing coefficients and relationship inference from parental pairs versus offspring.
• Utilized both single-point and multi-point genetic analysis approaches.

Main Results:

• For relationships closer than first cousin, parental pairs were more informative than offspring.
• Kinship coefficients from single-point methods provided accurate or equivalent genome sharing estimations for close relationships.
• For more remote relationships, offspring data was more informative, with inbreeding coefficients from multi-point methods yielding better estimations.

Conclusions:

• Parental pair and offspring data offer complementary information for relationship inference.
• Trio sampling is recommended when feasible, as it can improve the accuracy and range of inferred genetic relationships.
• The choice of sampling strategy and analytical method depends on the expected degree of relatedness.