Project 1: The Mouse Brain Library Project 2: Internet Microscopy (iScope) Project 3: Neurocartographer and Segmentation of the MBL Project 4: The Neurogenetics Tool Box


















Principal Investigator/Program Director Williams, Robert W.


Novelty of the Neurogenetics Tool Box

The novelty and utility of this project lie in its integration with the image libraries of the Mouse Brain Library, the data of the NeuroCartographer Project, and the genotypes provided by the Genotyping Core. Six innovative features will be integrated into the NTB over this 5-year grant cycle.

1.   Analysis methods not previously implemented. The fast interval mapping method (Whittaker, Thompson et al. 1996) has only recently been implemented in Map Manager QTX; it is not otherwise available in mapping software. The fast method for empirical significance thresholds (Peipho, personal communication) is not currently available in any mapping software.

2.   The integration of mapping software and genotype data. All current mapping programs come “empty.” A research group has to generate both genotypes and phenotypes. In contrast, the NTB will come loaded with a very large set of genotypes as well as with all the phenotypes automatically generated by the NeuroCartographer Project, which are simply a part of the MBL database (age, sex, body weight, brain weight, litter size, etc). We will have done much more than half of the data generation neuroscientists need for mapping QTLs. The NeuroCartographer Project will produce several hundred quantitative traits for each RI strain and for each animal in the advanced intercross. These values will be available to QTL mappers.

3.   Direct web access to QTL mapping tools. We intend to start with a simple and robust web interface in which investigators paste their phenotypes into a form with tab-delimited lists of values and then select or define trait names. For CNS phenotypes already in the NTB, users will be able to remap quickly to compare one trait with another. While web access to QTL mapping has been tried before (most notably in the Kearsey and Seatom QTL Café Project, 1997–1998, at the University of Birmingham), there is a major difference in our project. In our case, most of the data used for mapping will already be resident in the NTB. Neuroscientists will primarily be adding small numbers of traits. This fact gives us more control over the design of the

interface. For example, we know exactly how many cases there are, and we will know how to display their genotypes most effectively over the web. A prototype form for this purpose is illustrated and explained below.

4.   Correlative analysis of multiple traits. This is crucial for analysis of the CNS, in which different structures are often interconnected. For example, a group studying cell populations in the septal region will of course be very interested in knowing what genes control the size of the hippocampus. These types of correlated genetic studies have never been possible before because phenotype data have never been curated. This is even true for the RI strains. There is now no accessible database of the phenotypes of RI strains. In contrast, there are several sources for the genotypes of RI strains (B. Taylor, R. Elliott, R. Williams, and the Jackson Laboratory all maintain RI genotype databases). As part of the NTB, we will assemble phenotype databases for RI strains as well as for all of the advanced intercross animals and standard inbred strains in the MBL.

5.   Help in the early stages of QTL mapping. The neuroscientists who generate the trait values will often not have a strong statistics background. They will need help to evaluate the suitability of traits for QTL mapping. For example, we expect many neuroscientists to submit percentage and ratio data to the NTB: the ratio of the volume of the caudate nucleus to the substantia nigra, or the percentage of glial cells in the neocortex. These types of data will often need to be transformed and normalized prior to mapping. The NTB will include a set of web utilities that will screen and suggest how best to modify a data set prior to mapping.

6.   Extensive interactive online tutorials on QTL mapping. The statistical and bioinformatic aspects of QTL mapping need to be made more tractable for neuroscientists. Complex trait analysis should make a significant contribution to neuroscience. Successful penetration of this powerful method requires the right learning environment. A tutorial specifically tuned for neuroscientists is online at <>, and an extensive general tutorial is online at <>. These tutorials will be expanded and made significantly more attractive. We will even add short QuickTime movies using the expertise of the iScope project in video editing and QT movie production. We will step neuroscientists through the process of mapping QTLs using specific data sets from the MBL and the <> databases.



Next Topic