might not be dissected from the exactly same location and thus the brain tissue cell types could be slightly different, and (2) Set 1 samples consisted of both males and females while Set 1 samples were all males. Therefore, the biological differences between these two sets of samples may contribute more to the inconsistent results regardless of which platform was used. Similarly, the difference in cell types and proportions of different types of cells between hESC-derived neural cells and postmortem brain tissues is the most likely course why the in vitro cellular model study could not well confirm the findings from postmortem brain tissue studies. Moreover, we did not report AUD-associated and sex-specific transcriptomic changes when analyzing Set 1 samples because the subsample size (by sex) was too small to obtain unbiased results. In addition, we did not analyze other types of AUD-associated noncoding RNAs, such as long noncoding RNAs (lncRNAs). In the follow-up study, we will further analyze AUD-associated miRNA–mRNA–lncRNA regulatory networks.
