Table 1.

Evidence-Based Stages for Genetic Marker Development.

Stage	Goal	Types of markers potentially involved	Concerns	Marker-Assisted Selection (MAS)
1	Initial discovery of genetic association	Functional variants, markers in linkage disequilibrium (LD; inherited together) with functional variants, spurious associations	(A) Initial association in one animal set directs additional research. (B) Association could be spurious.	Not recommended
2A	Replication of association in validation animal sets	Functional variants, markers in LD with functional variants a	(A) Do markers have predictive value in the sense of consistent trait association? (B) Are there conditions under which predictive value might break down?	Supported by data for traits with replicated association, avoiding conditions where replication of association fails
2B	Assess potential correlated responses to selection	Functional variants, markers in LD with functional variants a	Are there trade-offs where some traits could be improved at the expense of other traits?	Conditions for use refined by assessment of potential trade-offs

^a While it is possible for spuriously associated markers to replicate association in separate animal sets, this is expected to be exponentially less likely with each test. Further, even in the unlikely event that such a marker did replicate association, repeated testing should quickly restrict the conditions under which use would be considered as the number of failed association attempts mount.