Agricultural Research Service
50th Anniversary

William M. Gelbart

E-mail

• Department of Molecular & Cellular Biology
  Harvard University
  16 Divinity Avenue
  Cambridge, MA 02138
  617 495-2906 (phone)
  617 496-1354 (fax)
  gelbart@morgan.harvard.edu

Faculty Positions

• Harvard University, Assistant Professor of Biology, 1976-1980.
• Harvard University, Associate Professor of Cellular and Developmental Biology, 1980-1983.
• Harvard University, Professor of Molecular and Cellular Biology, 1983-present.

Educational Background

• Harpur College, SUNY, Binghamton, NY, 1962-1963
• Brooklyn College, CUNY, Brooklyn, NY, 1963-1966; B.S., Biology, 1966
• University of Wisconsin, Madison, WI, 1966-1971; Ph.D., Genetics, 1971
• Predoctoral fellow with Dr. Allen S. Fox.
• California Inst. Tech., Pasadena, CA, 1971-1972; Developmental Genetics
• Postdoctoral fellow with Dr. Edward B. Lewis.
• University of Connecticut, Storrs, CT, 1972-1976; Genetics
• Postdoctoral fellow with Dr. Arthur Chovnick.

Current Major Departmental & University Activities

• Undergraduate and Predoctoral Education:
• Program Director, Interdepartmental Predoctoral Training Program in Genetics & Genomics, 2000-present.
• Head Tutor, Biology Undergraduate Concentration, 1994-present. (Director of the Biology Undergraduate Major in Biology)

Advisory

• Faculty Advisor, The Harvard Foundation, 1996-present. (Focal University organization devoted to multicultura l understanding).
• Member, University Committee on Research Policy, 2001-present.

National Scientific Advisory Committees

NIH Panels:

• Member, NIH Genetic Basis of Disease (GBD) Training Grant Study Section, 1984-1988; Chair, 1987-1988.
• Ad Hoc Member, Advisory Council to the National Institute of General Medical Sciences, 1/1989, 1/1994, 5/1996.
• Member, Human Genome Research Study Section, NIH, 1996-1999.
• Member, NHGRI Large-Scale Genome Sequencing Network Advisory Committee, 1999-present (Advisory Committee to Francis Collins, Director, NHGRI, on the human and model organism genome sequencing projects)
• Member, National Advisory Council to the National Human Genome Research Institute, Oct. 2000-2004.
• Chair, NHGRI Panel on Prioritization of New Genome Sequencing Projects, 2001-present.

Bioinformatics & Database Advisory Panels:

• Member, Genome Resources Advisory Committee, National Center for Biotechnology Information, National Library of Medicine, NIH, 1999-2000.
• Advisor, MGI - Mouse Genetics Informatics, Jackson Laboratories, Bar Harbor, ME, 1998, 2002.
• Advisor, WormBase, California Insitutes of Technology (C. elegans Genome and Genetic Database), 2001-present.
• Advisor, ZFIN - Zebrafish Genome and Genetic Database, U. Oregon, 2001-present.
• Member, Board of Directors,New England Complex Systems Insitute, Cambridge, MA, 1998-present.

American Cancer Society:

• Member, American Cancer Society Advisory Panel on Devel. Biology, 1992-1996.
• Ad Hoc Member, Extramural Advisory Council, American Cancer Society, 1999, 2001.

Consultant to Industry

• Consultant to Functional Genomics Therapeutic Area, Novartis Pharmaceuticals, 1999-present.
• Consultant to GeneLogic, Inc., 2000-2001.

Other Service and Honors

• Organizer, Fifty-fourth Annual Genetics Society of America Conference, 1985.
• Member, Board of Directors, Genetics Society of America, 1988-1990.
• Editor, Drosophila Information Service (volumes 69, 73, 74, 78, 79).
• Organizer, Thirty-third Annual Drosophila Research Conference, 1992.
• Fellow, American Association for the Advancement of Science, 1993.
• Recipient, NIH MERIT Award, 1995-2003.
• President, Drosophila Board of Directors, 1996-1997.

Research Program

Background: My long term interest has been in understanding the global interactions by which the gene products encoded by the genome mediate the cell-cell patterning events that underlie development. My laboratory has studied cell-cell patterning for the last 15-20 years, with the main emphasis on the identification and characterization of the gene products involved in TGF-beta ligand signaling in Drosophila (for example, we discovered decapentaplegic (dpp), the first member of the bone morphogenesis subfamily of the TGF-beta protein superfamily, co-discovered the first mutations in DPP receptors, and the discovered the Smad protein family, the central proteins in TGF-beta signal transduction (encoded by the Mothers against dpp (Mad) and Medea (Med) genes). Further, we have uncovered and did much of the initial characterization of the many of the fundamental patterning processes to which the DPP signaling pathway contributes and first recognized that DPP is likely to act in embryonic dorsoventral patterning and anteriori-posterior development of the adult appendage precursors (the imaginal disks) as a secreted, concentration-dependent morphogen. In addition to these studies, I have had a significant and long-standing interest in an epigenetic phenomenon, termed transvection, in which the expression of specific genes is modulated by the state of organization of the interphase nucleus.

Current Research Efforts: One of the exciting opportunities of the last few years has been to look at such problems globally through high-throughput experimental approaches and by computational analysis. In order to effectively carry out such analyses, it is crucial that there be a solid foundation of information on the structure and function of the genome and its encoded products. Building this foundation, especially in the major model system, the fruit fly Drosophila melanogaster, has become the central issue for my group for the last few years.

The finishing of the sequencing of the euchromatic portions of the D. melanogaster genome occurred about a year ago. Even with this full, high quality sequence to gaze at, understanding the nature of the products encrypted within that sequence and of the encrypted regulatory elements that control the deployment of those products remains challenging. Nonetheless, attaching as much molecular biological and phenotypic information to the genome’s sequences is of crucial importance to the success of efforts in the “post-genomic” or “functional genomic” era. Thus, my groups efforts are focused on several aspects of genome annotation.

Experimental Annotation of the Genome: Even in as well-studied a genetic system as D. melanogaster, our ability to efficiently mutate every known or predicted transcription unit in the genome is limited. We have developed a novel tool consisting of a compound transposable element containing two independent mobilization elements to aid in this process. Using this P{wHy} system molecularly defined nested sets of deletions can be generated bidirectionally from the insertion site of a single P{wHy} element residing at any location in the genome. The resulting deletions range in size from a few base pairs to several hundred kilobase pairs, and are generated at high efficiency. Within 50 kb of the starting insertion site, deletions can be generated with offset endpoints every kb or so, meaning that the likelihood of separating any two transcription units within 50 kb of a starting point is very high. We are saturating the genome for P{wHy} insertion sites as starting points for generating such deletions. (We estimate that already, 45% of the autosomal euchromatin is within 50kb of one of our inserted elements.) We are adapting the system for high throughput, and for use in such applications as (1) the efficient identification of the transcription units associated with each of about 1500-2000 anonymous recessive lethal/sterile complementation units (2) the homozygous deletion of anonymous transcription units by overlap of precisely-mapped deletions derived from P{wHy} elements bracketing the gene of interest, (3) in vivo mapping of extended regulatory elements, (4) high throughput preparation of RNAs for microarray/chip probes from mutant individuals such as those described in (2).

Computational Annotation of the Genome -- Gene Predictions: With the finishing of the euchromatic sequence of D. melanogaster, FlyBase has the responsibility to capture and synthesize all pubished information about the genome. (I am the PI of the FlyBase Consortium, with sites at Harvard, Berkeley (Gerry Rubin, co-PI), Indiana (Thom Kaufman & Kathy Matthews, co-PIs) and U. Cambridge (UK) (Michael Ashburner & Rachel Drysdale, co-PIs). One major aspect of this is to develop and maintain the authoritative predictions of the structures of all of the genes in the genome. Toward this end, FlyBase has recently swept through the finished sequence, producing a new canonical set of transcription unit structures for the 13,380 protein-coding genes. The reannotation effort has included use of experimental sequence data (cDNAs/ESTs), sequence similarity data, and ab initio gene predicitions. We are now following this up with more selective re-examination of the gene models based on newly available cDNA sequences and on comparisons with the sequence of a second Drosophila species -- Drosophila pseudoobscura -- and with two other insect genomes -- the malarial mosquito, Anopheles gambiae and the honey bee, Apis mellifera.

Computational Annotation of the Genome -- DNA Motifs: We are developing computer-assisted, interactive methods for identification of DNA motifs corresponding to regulatory elements and other protein-binding sites within the genome. Thus far, we have identified several different motifs at/near transcription start sites, including one that has properties suggestive of it being a novel initiator element. We are extending these studies to explore computational techniques for identifying motifs located remotely from transcription start sites. In addition, we are exploring predictions of non-protein coding genes, using small nucleolar RNAs (sno-RNAs) as a first testbed. We are complementing the computational studies with functional experiments to explore the validity and roles of the sequences thus identified.

Computational Analysis of the Genome -- Database Mining Techniques: FlyBase also captures a great deal of information of genes, alleles, molecules and phenotypes using a variety of controlled, hierarchically arranged vocabularies as well as free text. As part of our internal project, and also in collaboration with computer scientists at MITRE, Inc., we are continually exploring ways to mine information from data sets such as ours, to achieve interoperability with other biological databases, and to exploit natural language querying systems for the purposes of automatic data capture and data mining.

Computational Analysis of the Genome -- Quantitation of Anatomical Phenotype and Spatial Expression Patterns: While methods for analyzing molecules in a highly automated, high throughput fashion are either available or can be imagined, there is a pressing need for developing techniques to phenotype individuals with equally high efficiency. We have a small research project to develop such techniques and to apply these techniques to characterize the anatomical patterns and spatial expression patterns of wild-type and mutant individuals.