Over the last 5 years, the Human Genome Project has had an enormous impact on the scientific community, which is witnessing an explosive increase in the description of genes and disease-causing mutations. As a result of this overwhelming expansion of data, many problems have arisen in describing and cataloguing sequence alterations and making them accessible to researchers. Journals cannot publish all the mutation reports, and central databases such as Online Mendelian Inheritance in Man (OMIM) have found it necessary to limit the data they attempt to capture. Mutation information currently is fragmented and incomplete, and, for most genes, no database exists at all.

A few locus-specific databases (such as those for Factor IX, cystic fibrosis, PAH, and BRCA1) have developed from the need of individuals or consortiums to keep track of data for their research. However, the usefulness of these smaller databases is limited by their lack of uniformity in design, content, and nomenclature making access, communication, and analysis difficult.

This scarcity of up-to-date gene-mutation listings hampers researchers and clinicians in determining whether a particular mutation has been described or not. Also, investigators lack comprehensive information on the loss of function of different mutations in specific genes, and clinicians are unable to draw on experiences of others who have patients with similar mutations.

Mutation Lists Vital

Complete and current mutation lists are vital for studying genotype-phenotype relationships, treating patients with similar phenotypes, and analyzing amino-acid residues important in the function of gene products. Mutation lists also are needed to verify the existence of modifier genes and assist in their identification, monitor mutagenic environmental influences in somatic oncogenes and tumor-suppressor genes, design diagnostic protocols, assist in manuscript review, and examine the spectrum and type of gene mutations. Although rare, disease-associated mutations are powerful probes of genetic variations in populations. Knowledge of these mutations allows researchers to provide information and services to those affected and to characterize mutation effects in populations.

Because researchers believe mutations in the human genome are likely to number in the millions, a systematic approach to collecting and maintaining mutation data is needed. Some progress toward that goal has been achieved: Some researchers have made their databases available via the Internet (e.g., phenylketonuria, cystic fibrosis, mutations in factor IX, and P53), and central databases such as OMIM have compiled partial listings of mutations identified in specific genes. The Human Gene Mutation Database of David Cooper and Michael Krawczak (Institute of Medical Genetics, Cardiff, Wales) contains mutations and an index of where they are published (http://archive.uwcm.ac.uk/uwcm/mg/hgmd0.html).

Database Association

To make comprehensive mutation lists available for research, investigators are now forming an association of curator-driven, locus-specific databases whose standardized content can be downloaded onto central databases. Many believe these databases can serve the community better because they usually are set up and maintained by researchers, contain numerous unpublished mutations, have a greater likelihood of being complete, and avoid the limitations of central databases. Their bulletin-board components contain useful information on primers, methods, and scanning laboratories. A directory of locus-specific databases will be added to the Web site. Another directory comprises listings in OMIM.

Project History

At an October 1994 meeting in Montreal, called to consider nomenclature, further discussion led to an agreement to work toward placing up-to-date mutation lists on the Internet. Attendees felt that the curator-driven system should be simple and contain both published and unpublished mutations. This approach was supported by Victor McKusick (OMIM), Charles Scriver (McGill University, Montreal), Haig Kazazian (University of Pennsylvania), Lap-Chee Tsui (Hospital for Sick Children, Toronto), Aravinda Chakravarti (Case Western Reserve University), and Douglas Wallace (Emory University) as well as the American Society of Human Genetics (ASHG), Human Genome Organisation (HUGO), and March of Dimes (MOD).

A formal meeting in Minneapolis in October 1995, supported by MOD and organized by HUGO, included systems experts and representatives from prominent mutation databases, genetic and cancer societies and journals, publishers, and central databases. Attendees endorsed the idea of an alliance of locus-specific database curators who would keep their own databases current and assist in implementing a unified and uniform approach. Working groups were established on Software and Content (Chair, Scriver), Central Database [Chairs, McKusick and Jim Ostell (NCBI)], Nomenclature [Chair, Stylianos Antonarakis (University of Geneva)], and Mutation Database Association (Chair, Richard Cotton).

At a satellite meeting of HGM '96, Michael Ashburner (University of Cambridge) outlined a proposal for a new central mutation database that would supplement individual mutation databases. The central database would contain core information rather than locus-specific data and allow faster analysis of stored data. In preparation for a half-day meeting to be held October 29 in conjunction with the ASHG meeting in San Francisco, HGM '96 attendees agreed to place the working groups' discussion documents on the Web site and to initiate a mutation electronic mailing list. The mutation-database project is now a HUGO initiative and - collaboration among Scriver, McKusick, and Cotton. (Chair).