johnny,
The nature of the entire technology is the attempt to directly connect the genetically influenced characteristics of individuals to the genetic variances that cause them. If a difference in the code of someones DNA causes a specific disease or susceptibility, then the ideal analysis will discover that exact variance and connect it directly to the resultant trait.
Such a direct connection between genetic cause and effect is only the result of "genotype- or haplotype-based association analyses". Admixture mapping CANNOT make such a connection.
Admixture mapping can only indicate a very vague probability regarding a trait and its variance. For instance suppose a genetic variance entered the human genome somewhere in Europe 30 thousand years ago as a result of a genetic mutation in a specific individual(Call him Alpha). Every person today that carried that specific variant would be a direct descendant of that individual. Everyone would also carry a portion of the same set of AIMs as each other. Those AIMs might be considered an alternative set of search criteria to the trait in question, until you realize that every descendant of Alpha's brother (who don't carry the variant) have exactly the same set of AIMs. Not to mention everyone else in the village/region at the time.
AIMs are a very crude and imprecise method of determining the connection between variant and trait. Everyone wants to know the 'direct' connection and that is why they all prefer direct genotype or haplotype-based association analysis.
Admixture mapping may present some interesting trends, but it does not provide the cause and effect relationships that are essential to the eventual understanding of the genome.
regards,
frog
AI
Market Data 
Markets 
