Number of loci required for the likelihood-ratio method to achieve the same α and β as 100 loci using an exclusion-based method

Embedded ImageEmbedded ImageEmbedded ImageEmbedded Image
Embedded Image697780697780BSi107113114556061
Embedded Image738083718082BSe144142141242221
Embedded Image778587768386HU778486728184
Embedded Image768486728184HDFC828991788688
Embedded Image828991788688HSi899596838991
Embedded Image899597838991HSe111113114788588
Embedded Image889294637680D1151158161344143
Embedded Image96101102718184D2100100100576264
Embedded Image110113113788587D389a96a98a697779
  • The T column denotes the true relationship of the trio. For each relationship, there are six columns. The first three columns give the number of loci required to have α and β comparable to the exclusion method when using the test statistic Embedded Image and when q = 0.2, 0.35, and 0.5 as indicated by the column headings. The fourth through sixth columns to the right of each relationship show the number of loci required for a test of relationship T vs. relationship Q, based on Embedded Image (the most powerful test statistic that could be used if all the trios were known a priori to be either of type Q or of type T), to have the same α and β as the exclusion-based method.

  • a Values that must be treated cautiously because the accuracy of the importance-sampling methods is questionable for these relationships.