Category fluency, latent semantic analysis and schizophrenia: a candidate gene approach

Area of Science:

• Neuroscience
• Genetics
• Computational Linguistics

Background:

• Category fluency tasks assess neurocognitive function and are sensitive to cortical dysfunction, particularly in schizophrenia.
• Traditional metrics include word count and error rates (non-category words).
• This study integrates computational linguistics with genetics to investigate the genetic architecture of fluency measures.

Purpose of the Study:

• To examine the genetic architecture of category fluency using both traditional and novel computational linguistic approaches.
• To explore the relationship between genetic variations and cognitive processes involved in verbal fluency, learning, and recall.
• To identify specific genes associated with category fluency performance in a schizophrenia cohort.

Main Methods:

• Applied Latent Semantic Analysis (LSA) to analyze semantic clustering of generated words, reflecting memory search.
• Included standard verbal learning and recall tests (Wechsler Memory Scale, CVLT).
• Utilized a candidate gene approach, analyzing 39 coding SNPs from a schizophrenia GWAS, focusing on genes related to language, learning, and processing speed in 665 subjects.

Main Results:

• Computational analysis of word content provided neurocognitively viable metrics beyond traditional fluency measures.
• Identified and discussed three replicated Single Nucleotide Polymorphisms (SNPs) in ZNF804A, DISC1, and KIAA0319 genes.
• Preliminary findings suggest a genetic component to category fluency performance, with potential implications for understanding schizophrenia.

Conclusions:

• Extracting semantic meaning from verbal fluency responses yields valuable neurocognitive metrics.
• Specific genes (ZNF804A, DISC1, KIAA0319) show potential associations with category fluency.
• Computational analysis of linguistic data holds promise for future genomics research in cognitive disorders.