Each content category was submitted to Affymetrix as a list of markers or genomic regions. Affymetrix used proprietary software and a Axiom-validated marker database to select the best-performing markers (one or more probesets per marker) that covered the targeted content, either through direct inclusion or through efficient pairwise tagging. Multiple probesets were selected for markers that are either non-validated or deemed high priority (e.g., markers with known associations with addiction, smoking behavior, or nicotine metabolism), in order to minimize genotyping failures of these markers.
