Guillermo F Poblete1, Savannah N Gosnell2, Matthew Meyer3, Mary Fang4, Tien Nguyen5, Michelle A Patriquin6, David Nielsen7, Thomas Kosten2, Ramiro Salas8. 1. Universidad de Buenos Aires, Facultad de Ciencias Exactas, Buenos Aires, Argentina; The Menninger Clinic, Houston, TX, USA. 2. Menninger Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine, Houston, TX, USA; Michael E DeBakey VA Medical Center, Houston, TX, USA; Department of Neuroscience, Baylor College of Medicine, Houston, TX, USA. 3. Baylor College of Medicine, Houston, TX, USA. 4. Baylor College of Medicine, Houston, TX, USA; The University of Texas at Austin, Austin, TX, USA. 5. Rice University, Houston, TX, USA. 6. The Menninger Clinic, Houston, TX, USA; Menninger Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine, Houston, TX, USA; Michael E DeBakey VA Medical Center, Houston, TX, USA. 7. Menninger Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine, Houston, TX, USA; Michael E DeBakey VA Medical Center, Houston, TX, USA. 8. The Menninger Clinic, Houston, TX, USA; Menninger Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine, Houston, TX, USA; Michael E DeBakey VA Medical Center, Houston, TX, USA; Department of Neuroscience, Baylor College of Medicine, Houston, TX, USA. Electronic address: rsalas@bcm.edu.
Abstract
BACKGROUND: Linking human genetics and brain imaging data is extremely challenging, among other reasons because both fields suffer from multiple comparison problems. NEW METHOD: ProcessGeneLists (PGL) links genetics and human brain imaging by using genes associated with a disease and calculating a normalized mRNA expression average of those genes in each brain region. Brain regions in which those genes are most co-expressed become regions of interest (ROIs) to perform brain imaging in participants with and without the disease, decreasing multiple comparisons. Once a region is identified as "imaging-relevant", the genes most responsible for that ROI being highlighted can be genotyped in the imaged sample. This allows to re-analyze imaging data under the light of likely relevant genetics, to study possible brain imaging/gene variant interactions. RESULTS: As proof-of-concept, we created two lists of genes expressed in the habenula and the striatum, to verified that PGL would highlight those regions. Next, we used a list of genes likely important in alcohol abuse from the literature, which identified several brain regions previously associated with alcohol abuse such as the striatum, habenula, and hippocampus. COMPARISON WITH EXISTING METHODS: To our knowledge there is no current method to obtain brain regions of interest from genetics data. CONCLUSIONS: Genetics typically asks "which genes are associated with a disease?" while human brain imaging typically asks "which brain regions are associated with a disease?" PGL asks "which genes, via modulation within specific brain regions, are found to be associated with a disease?".
BACKGROUND: Linking human genetics and brain imaging data is extremely challenging, among other reasons because both fields suffer from multiple comparison problems. NEW METHOD: ProcessGeneLists (PGL) links genetics and human brain imaging by using genes associated with a disease and calculating a normalized mRNA expression average of those genes in each brain region. Brain regions in which those genes are most co-expressed become regions of interest (ROIs) to perform brain imaging in participants with and without the disease, decreasing multiple comparisons. Once a region is identified as "imaging-relevant", the genes most responsible for that ROI being highlighted can be genotyped in the imaged sample. This allows to re-analyze imaging data under the light of likely relevant genetics, to study possible brain imaging/gene variant interactions. RESULTS: As proof-of-concept, we created two lists of genes expressed in the habenula and the striatum, to verified that PGL would highlight those regions. Next, we used a list of genes likely important in alcohol abuse from the literature, which identified several brain regions previously associated with alcohol abuse such as the striatum, habenula, and hippocampus. COMPARISON WITH EXISTING METHODS: To our knowledge there is no current method to obtain brain regions of interest from genetics data. CONCLUSIONS: Genetics typically asks "which genes are associated with a disease?" while human brain imaging typically asks "which brain regions are associated with a disease?" PGL asks "which genes, via modulation within specific brain regions, are found to be associated with a disease?".
Authors: Guillermo Poblete; Tien Nguyen; Savannah Gosnell; Olutayo Sofela; Michelle Patriquin; Sanjay J Mathew; Alan Swann; David A Nielsen; Thomas R Kosten; Ramiro Salas Journal: Chronic Stress (Thousand Oaks) Date: 2022-03-21