Summary

Project Participants:

David Lightfoot, Ph.D., Southern Illinois University, Carbondale,

Khalid Meksem, Ph.D., Southern Illinois University, Carbondale,

Christopher Town, Ph. D., The Institute for Genomic Research,

Hongbin Zhang, Ph.D.Texas A&M University

Introduction:

Functional genomics requires multi-parallel approaches to assess gene function that make use of the information and reagents provided by structural genomics. Genomics has enhanced understanding of the nature of complex trait loci through physical maps and through expression profiles that underlie phenotypes. Integrating tools that allow gene targeting, (STS, ESTs, genome sequence tags and DNA markers) with physical maps has lead to the routine isolation of DNA fragments that contain candidate genes underlying complex trait loci.

Proposed is the development of a high-density gene map from a robust physical map of the soybean genome. The integrated map will allow us to test the hypothesis that ‘in the plant post-genome sequence era, crop whole genome sequencing is unnecessary for functional genomes’. Testing the hypothesis will produce a map for functional genomics that will accelerate the rate of economically important gene discovery; provide a framework for integrating functional gene maps of legumes; and provide a tool for comparative structural and functional genomics among crops.

Objectives and Goals: We will:

1. Develop at high-resolution a 40-60 thousand gene map that can be used for functional genomics. Comparison of the gene map with the predicted gene content of 9 Mbp of genome sequence will test the hypothesis that whole crop genome shotgun sequencing is redundant.

2. Improve methods for assembling the fingerprinted BACs into complete and robust physical map contigs. Comparisons of physical map construction methods will address the validity of gene locations inferred from physical maps.

3. Sequence 8-9 Mbp of soybean genomic DNA from a gene rich region on chromosome G. The contig generation tools and gene maps can be critically evaluated compared to benchmark sequence data for overlapping BACs and their predicted gene content.

4. Test the utility of the contigs, virtual sequence and DNA sequence for identifying homeologous genomic regions. The predictive value of synteny associations will be tested.

5. Provide researchers with electronic access to fingerprints, maps, genome sequences and clones. Allow interactive analysis of regions likely to contain genes and QTLs of agricultural importance and determine the robustness of the virtual sequence represented by the gene map.

Methods:

The primary method for gene map development and physical map gap filling will be STS integration with BAC fingerprint data. Paralogs will be identified from ESTs to increase the map’s gene density. DNA sequencing will be used to generate 8-9 Mbp of sequence from two gene rich regions and one soybean chromosome. Paralogs and sequence predicted genes will be annotated by identification in the transcriptome, hypo-methylated genomic DNA or DNA derived from hyper-acetylated nucleosomes. Bio-informatics will be used to compare gene order across genera and to make gene function predictions.