cohort follow-up samples. Power calculations that document the modest power in European-Americans samples revealed even more modest power for the smaller number of African-American substance dependent individuals in this sample; we have thus not analyzed these samples. Modest power limits interpretation of negative data, substantially restricting inferences about genes identified in the more robust dataset from MNB research volunteers but not in these ECA samples. 3) There is very highly significant confidence in the overall set of convergent positive results reported here. However, the values for each gene, tested individually, provide much more modest levels of statistical assurance. 4) Focus on data from autosomes here allows us to combine data from male and female subjects, but misses potentially important contributions from sex chromosomes. 5) The individuals in the Baltimore ECA cohort were not initially sampled based on their willingness to be volunteers. However, participants needed to consent in order to be able to be followed and studied genetically. Although the overwhelming majority of the European-American participants who were followed did consent to participation in genetic studies, potential contributions that the non-consenting individuals might have made to the present results remain unknown. 6) The pooling approach that we use here provides
