Multifactoriality in Psychiatric Disorders: A Computational Study of Schizophrenia

Area of Science:

• Neuroscience
• Computational Psychiatry
• Systems Biology

Background:

• Limited success in identifying biological causes for mental disorders.
• Growing dissatisfaction with current diagnostic classifications.
• Questionable assumption of specific microscale brain alterations for clinical syndromes.

Purpose of the Study:

• Evaluate the multifactoriality of alterations in psychiatric illness.
• Explore computational models of schizophrenia.
• Investigate the diversity of causes for schizophrenia-related phenotypes.

Main Methods:

• Parametric exploration of published computational models of schizophrenia.
• Simultaneous variation of multiple parameters (receptor conductances, connectivity, excitation).
• Analysis of network models for schizophrenia symptoms and spatial working memory deficits.

Main Results:

• 5.1% of generated networks exhibited schizophrenia-like symptoms.
• 4.9% of networks showed schizophrenia-like alterations in spatial working memory models.
• Individual parameters lacked sensitivity and specificity in identifying schizophrenia-like networks.

Conclusions:

• Schizophrenia-related network phenotypes can result from numerous molecular and circuit-level causes.
• Unified genetic or molecular explanations for the full syndrome or its endophenotypes are unlikely.
• Multifactoriality is a key consideration in understanding psychiatric disorders.