rather, there is an interest in re-organizing the nosology or groupings of mental disorders to reflect causes rather than symptoms.
For the most part, mental illnesses have over the history of psychiatry been labeled and placed into groups simply on the basis of somewhat shared symptoms, rather than underlying mechanisms of causation. This was unavoidable as underlying mechanisms weren't (and actually remain) uncertain or unknown.
The result of that historic approach was construction of over a dozen groups of mental illness such as developmental disorders, substance abuse disorders, mood disorders, somataform disorders, eating disorders, impulse control disorders, personality disorders, etc.
In the last few years, through the application of molecular biology to mental illness it has become apparent that some disorders placed in different categories used in the symptom based DSM actually are associated with the same gene variations.
The basic idea is to form a new taxonomy of mental disorders that uses genetics and underlying shared mechanism to group mental illness into categories that are biologically similar.
The advantage of a such a system would be that biologically similar disorders would likely be caused by problems in similar mechanisms and the disorders would respond to similar classes/types of interventions.
The paradigm that NIMH is moving to envisions mental illness as a hierarchical association of mechanisms. This system begins with genetics endowment, moves through the expression of genes that influence various biological systems that intern contribute to or influence neural modules/function related 'brain circuits' that affect various recognized psychological domains (Cognition, emotions, attractions and aversions, mood etc.).
Within that there is no reason to doubt that a recognizable disorder such as depression will still exist. But what may happen is that it becomes grouped with currently seemingly unrelated illness because of shared underlying biological mechanisms.
It seems unlikely that the new groupings will look very much like the existing groupings--partly because scientists like to leave fingerprints in their disciplines, and nothing does that so much as a reordering of the subcategories considered by a discipline.
However, in a curious in not ironic turn of events, the new domains for research were formed around various groups of symptoms/systems affected by psychopathological processes: Negative Valence Systems, Positive Valence Systems, Cognitive Systems, Systems for Social Processes and Arousal/Modulatory Systems.