Second, when dealing with peripheral blood cells, it is important to note that differences in cell mixture may confound differences in methylation patterns due to substance use. Since this problem has been recognized, it has become routine to control for such differences in cell composition through either direct cell counts or through the method of Houseman and colleagues [26], by which methylation patterns in a given sample can be analyzed to infer the original cell mixture distribution. However, it is important to note that the effect of these measures in improving signal in some disorders is debatable [27,28,29] and the recent discovery that many lymphoid cells do not display traditional cell specific markers used to develop the methylation data used in the Houseman technique suggests a need to refine these cell correction approaches [30].
