Browsing by Subject "Neuroscience methods"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Methods to minimize subjectivity in correlating brain structures to phenotypes-of-interest
    (2020-05) Fang, Mary; Marinelli, Michela
    Mapping the brain structures that underlie a phenotype-of-interest (POI) has implications in explaining, diagnosing, and treating pathophenotypes. There are many methods that assess the correlation between candidate brain regions and POIs. One of these methods involves experimentally manipulating the candidate brain region in animals and studying the effect on a POI. Another common method involves brain-imaging studies, which compare brain images of cases (individuals with the POI) versus controls (individuals lacking the POI). Both of these methods face the limitation of ambiguity in determining effect size in the correlation between structure and function. For experimentally manipulated brains, there is not yet an objective, standard way to describe the size and location of the region of manipulation (ROM). Scientists in the field currently describe the size and location of the ROM based on a qualitative, ordinal scale. Similarly, case-control neuroimaging studies are prone to inflated false positives. This issue is further compounded by the limited sample size due to the high cost of brain imaging. Without objectively defining the brain regions that contribute to a POI, it is challenging to accurately assess the relationship between structure and function. Two novel methods are discussed, each with the aim of decreasing subjectivity and ambiguity in identifying brain regions that play a role in a POI: (1) quantification-of-ROM, a method to quantify the size of the ROM relative to a color-coded reference brain atlas and (2) ProcessGenesList (PGL), a computational method that combines genetics with neuroimaging to filter for brain regions of interest (ROIs). Preliminary results of Quantification-of-ROM were obtained and quantified the ROM across 6 brain regions in 5 different brain slices in a single brain. A pilot study was conducted as a proof-of-concept for PGL. Genes known to be differentially expressed in the habenula comprised a GeneList. This GeneList was run through PGL, which correctly identified the habenula as an ROI. Both of these methods achieve similar aims in novel ways, and similarly, will each have potential utility across neuroscience research and medicine. Quantification-of-ROM is useful for testing hypotheses of whether a brain region is important for a POI and in elucidating the role of an understudied brain region. One application of the quantification-of-ROM method is to assess the role of an understudied brain region, the lateral preoptic area (LPO), on reward-seeking behavior. PGL can also serve as a hypothesis-generating tool by narrowing down ROIs to be studied for a particular POI.