Multi-dimensional genetic analysis in the BXD recombinant inbred (RI) genetic reference population. QTL mapping of a factor predictive of naloxone-induced morphine withdrawal, naloxone (30 mg/kg i.p.) after morphine (50 mg/kg dose i.p.). Blue traces are likelihood ratio statistics. The red/green trace plots the additive effect with green representing increase or effects of DBA/2J alleles and red representing positive effects of C57BL/6J alleles. Horizontal lines represent the p < 0.05 (red) and p < 0.63 (blue) empirical significance thresholds. Analysis of 3 distinct measures of morphine withdrawal reveals no significant QTL, although suggestive loci are present. The combined analysis of multiple behaviors through a reference population reveals a significant QTL for a derived factor of these and other correlated behaviors [22]. The “Morphine Withdrawal Factor” was derived by performing a maximum likelihood factor analysis on measures from our high throughput behavioral study in the BXD RI lines

Convergent evidence for poorly studied genes associated with alcoholism. This network was constructed from 31 alcohol-related gene sets based on 3 major experimental data types: (i) QTL candidate genes, (ii) GWAS candidates, and (iii) differentially expressed genes from microarray experiments. These data came from 5 species (fruit fly, zerbrafish, mouse, monkey, and human). At the top of the tree are nodes representing multi-way intersections of the gene sets, and at the bottom are the individual genesets. The map of all gene set intersections contained more than 100 nodes. Bootstrapping was applied to reduce the complexity. Purple shading indicates nodes that contain genes previously associated with human alcoholism or mouse alcohol-related phenotypes. This analysis reveals that although many genes have been identified in specific studies of alcoholism, the most highly represented genes among the 31 data sets are not currently annotated to alcoholism

A schematic representation of the integrative multi-dimensional analysis of behavior and the relation of substance-use phenotypes to previously established factors of behavior in a reference population

An overlay of representative QTL scans centered at their peak loci mapped with approximately 250 to 300 mice in each study. The entire chromosome containing each representative locus is shown. QTL precision is dramatically improved with new reference populations, including the Collaborative Cross, Diversity Outbred, and to a lesser extent, the expanded BXD recombinant inbred panel. A panel of consomics mice has the ability to narrow a QTL to the resolution of a single chromosome [30]. Thirty-five BXD recombinant inbred strains were used to map basolateral amygdala complex size, resulting in a significant QTL on chromosome 8 [41]. The precision of an F2 cross produces a small, broad significant QTL for bone mineral density on chromosome 7 [42]. From the expanded BXD RI population, a narrow QTL was mapped on chromosome 13 for open field rearing after cocaine [22]. Using mice from the collaborative cross breeding population a QTL for liter size was mapped to a narrow locus on chromosome 6 [14]. Early studies using the Diversity Outcross population of mice reveal high precision loci for light-dark box behavior. CC = Collaborative Cross; DO = Diversity Outbred; F2 = F2 hybrid cross; LOD = Logarithm of odds; MBp = megabase pairs