6.916667
#50
POPULATION BASED CHARACTERISATION OF HLA TYPING KITS AND AMBIGUOUS TYPING RESULTS: THE FREQUENCY INFERRED TYPING (FIT) INDEX.
Wolfgang Helmberg PhD 3, Douglas Hoffman 1, Michael Feolo 1 and Martin Maiers 2. 1 NCBI, NIH, Bethesda, MD, USA ; 2 NMDP, Minneapolis, MN, USA and 3 UBT, Med.Univ Graz, Graz, Austria .
HLA probe based typing usually leads to an ambiguous list of allele combinations. We present the population based FIT index that describes the probability of a correct allele assignment for various ambiguous typing scenarii.
The FIT index is calculated by comparing the product of the allele frequencies of the allele pair selected with the sum of the products of all other allele pairs that share the same typing pattern. It represents the negative logarithm of the probability that the picked allele combination is incorrect within a distinct population (FIT 3: 1 in 1000, FIT 4: 1 in10000).
The FIT index can be applied in a number of settings:
1) The characterisation of an allele call in the context of an ambiguous typing result.
2) The description of the resolution of individual typing kits. This provides an easy tool to compare and improve typing kits or determine which typing kit is best suited for a distinct population. The FIT index can be used to characterise a kit on the generic and the allele group specific level.
3)The Probability of finding a matching donor who has been typed just at low resolution.
4) It allows to define upfront the required level of resolution. E.g. registry typing of new donors (FIT>=3) versus clinical typing of a donor going to transplant (FIT >=6).
dbMHC at the NCBI has generated FIT indices for all typing kits that have been submitted to its typing reagent database. FIT indices for generic and allele group specific typing are available for the 7 main population areas of the IHWG anthropology project and the 5 NMDP population groups. URL: www.ncbi.nih.gov/mhc, section typing kits.