I n t h i s i s s u e ...Brain biochemistry is rich and complex, and brain func-tions are also varied, complex, and interdependent. Drugsfor treating central nervous system (CNS) diseases act onthe biochemical machinery of the brain, on cell mem-brane or nucleus receptors, on second messengers, andon neurotransmitters. This leads to changes in macro-scopically measurable brain functions such as perception,motor activity, memory, emotion, consciousness, andmany other functions that enable our adaptation to ourchanging internal, physical, and social environments. The first CNS drug discoveries during the 1950s and1960s were serendipitous to some extent, but also guid-ed by astute hypotheses. At that time, a series of usefuldrugs could be developed by one researcher, or by arather small team, which has become near to impossiblenow. Several drugs that were discovered in those timesremain standards, to be used as comparatives in the cur-rent evaluation of the clinical efficacy of drugs in devel-opment. Researchers in pharmaceutical sciences and industriesnow have a huge knowledge base in fundamental andclinical neurosciences, from which to develop innovativedrug treatments for neurological or psychiatric diseasesand disorders. Indeed, nowadays, a pharmacological com-pound is described in an exhaustive manner, covering allits influences (by analogy with an antibiogram): ie, areceptorgram, a transportergram, an enzymogram, agenogram, a proteinogram, a metabogram; these termsconcern the list of all biochemical constituents that arechanged during drug administration. To this biochemicallevel of description, one can now add a physiological level,ie, the drugs influences on neuronal networks devoted togiven CNS functions. Neuroimaging technologies are rel-evant tools at this level of analysis. As an illustration of aphysiological description of pharmacodynamics, someantidepressants influence the automatic and noncon-scious perception of others emotions, by dampering theactivity of the amygdala and related structures. The finalgoal of neurologic or psychiatric drug treatment is thecontrol of symptoms, acutely, chronically, or with a pre-ventive goal, ie, to act in a bottom-up manner, from mol-ecules to mind.In this issue of Dialogues in Clinical Neuroscience, theauthors describe how researchers look for new therapeu-tic compounds, ie, how they explore unknown domainsto discover those structure/activity relations that couldprovide therapeutic effects with minimal side effects. Anoverall conclusion to be drawn from these articles is thatthe challenge of new drug discovery is immense, andrequires intelligence and the use of shortcuts, as well as athorough and cautious approach. To these research com-petences, one should also add real administrative talent.A group of lazy men once returned after a terrible day'swork: there was work for at least six people, but fortu-nately there were twelve of them! Looking at the contri-butions by the authors of this issue, I am persuaded thatthe reverse situation prevails in the domain of drug dis-covery: a small number of people have to achieve a lot. A State of the art paper on the expression profiling ofdrug response, from genes to pathways, by Ralf Herwigand Hans Lehrach, provides a relevant introduction to thetheme of drug discovery and proof of concept. Indeed,the domain of therapeutics will benefit from innovativeideas on how to identify targets of medication. The recentfunctional genomic technology with DNA arrays ormicroarrays, ie, the capacity to measure the transcriptionof thousands of genes into RNA molecules, enables us tostudy the configuration of genomic changes induced bya given drug treatment. This advance offers a realisticapproach to pharmacology, in the sense that the phar-macodynamics of drugs is now addressed at its actuallevel of complexity, taking into account, at least theoreti-cally, all changes in protein synthesis. The authors describethe DNA transcription techniques and then review the dif-ficulties and the pitfalls of these techniques at several lev-els, ie, in the laboratory, in data analysis, in statistics, instandardization, and in the necessary international col-laboration between institutions. The authors then givetheir view as to how many new drugs could be found onthe basis of well-defined molecular targets; they alsomention avenues for the integration of functionalgenomics techniques with other techniques, with the aimof understanding the successive steps from DNA trans-duction to neuronal metabolism, to neuronal networksand then to global physiological systems, analyzed macro-scopically. Functional genomic technology can lead, notonly to new medications in neurology and psychiatry, butalso to a more efficient prediction of the efficacy and safe-ty of a given medication for a given patient. In his Pharmacological aspects paper, Michael Speddingdiscusses new directions for drug discovery. First, he