Toward Robust Functional Neuroimaging Genetics of Cognition

Area of Science:

• Cognitive Neuroscience
• Neuroimaging Genetics

Background:

• A common assumption posits that brain function measures are closer to biology than behavior, offering greater potential for uncovering genetic pathways.
• Influential studies using functional magnetic resonance imaging (fMRI) have reported associations between common DNA variants and altered neural activation in language regions.

Purpose of the Study:

• To directly test claims from a prior influential fMRI study (Pinel et al., 2012) on the genetic basis of neural activation in language processing.
• To investigate associations between specific DNA variants (SNPs) and brain activation in language-related regions using a replication approach.

Main Methods:

• Employed a closely matched neuroimaging genetics approach in independent cohorts.
• Utilized four times the sample size of the original study (427 participants vs. 94).
• Performed formal Bayesian analyses to assess evidence for or against the null hypothesis.

Main Results:

• No replication of the previously reported associations between specific DNA variants (SNPs in FOXP2 and KIAA0319/TTRAP/THEM2) and neural activation patterns.
• Bayesian analyses provided substantial to strong evidence supporting the null hypothesis (no effect).
• The study demonstrated sufficient power to detect smaller effect sizes than the original report.

Conclusions:

• The findings strongly refute the original claims regarding genetic associations with task-based fMRI activation in language regions.
• Highlights potential for elevated false-positive rates in functional neuroimaging genetics studies with small sample sizes.
• Advocates for large sample sizes, power calculations, and independent cohort validation for reliable identification of true biological signals in neuroimaging genetics.