In this short report, we address some practical problems in performing like
lihood-based allelic association analysis of case-control data. Model-free
statistics are proposed and their properties assessed by simulation, and pr
ocedures based on permutation tests a re described for marker-marker as wel
l as marker-disease associations. A memory-efficient algorithm is developed
which enables several highly polymorphic markers to be analysed. Copyright
(C) 1999 S. Karger AG, Basel.