Collection of variation causing disease - The Human Variome Project

Collection of mutations (defined as variation causing disease in this paper) causing disease began soon after the cause of thalassaemia as a mutation in the β-globin gene was established [1]. As other disease genes were defined, collection of individual mutations and their effects increased dramatically and it is still increasing today. The reason for this collection was/is for research, clinical guidance and diagnostic strategies in specific diseases. These collections have been made by expert and dedicated workers world- and disease-wide for their own purposes. Some collection has been occurring centrally, with mutations in all genes involved; for example, Online Mendelian Inheritance in Man (OMIM; http://www.ncbi.nlm.nih.gov/omim/[2]), Human Gene Mutation Database (HGMD; http://www.hgmd.cf.ac.uk/[3]) and MutDB (http://mutdb.org/[4]).

Collection of mutation data in individual genes into databases (locus-specific databases [LSDBs]) is difficult to fund and is rarely funded [5]. Thus, the fact that so many exist[6] indicates they are needed. These 'curators' consequently usually work in isolation in their spare time, using a variety of software packages, with data consequently being out of date, or, sometimes, LSDBs are removed from the internet. Other problems besides these are lack of incentives to submit data, lack of standard methods, lack of coordination, large amounts of data in clinics not available to those who need it and lack of secure repositories, among others.

Attempts to raise the profile of this activity and its funding began in the mid-1990s with the formation of the HUGO-Mutation Database Initiative,[7] which developed into the Human Genome Variation Society (HGVS; http://www.hgvs.org/) This organisation developed standards and provided advice and some incentives, and generated more LSDBs. To raise the profile further, the concept of the Human Variome Project (HVP; http://www.humanvariomeproject.org/[8])was developed. A meeting, co-sponsored by the World Health Organization and the American College of Medical Genetics (ACMG), was held in Melbourne, involving the United Nations Educational, Scientific and Cultural Organisation (UNESCO), Organisation for Economic Co-operation and Development (OECD), top genetics organisations and representatives from over 30 countries [9, 10]. The 96 recommendations published[10] indicated the poor state of the field. A further meeting was held in May 2008 and the plans developed from this were recently published [11]. Both these meetings and fora organised by HVP, have stimulated much useful activity (see below).

In general terms, it is clear that the genomics/genetics community is working towards not only annotating regulatory regimes and other features in the Encyclopedia of DNA Elements (ENCODE) project[12] (http://www.genome.gov/10005107), but also, ultimately, in five to ten years, towards having a standard reference sequence numbered from 1 to 3×10⁹. Each base may have a significance, and this significance will be accessed by clicking on the base. The single nucleotide polymorphism, HapMap, 1000 Genomes Project and copy number variations (CNV) consortia will provide many variations and some will be linked to common disease or pharmacological utility. It is important to realise that the vast majority have no effect on humans or will have no recorded phenotype. On the other hand, data derived from inherited disease will have phenotype and will be useful not only in the clinical sphere, but also in the 90 per cent in general of human disorder outside the 10 per cent (average) of all diseases inherited in a Mendelian manner. The number of mutations causing inherited disease, at least within or near the coding sequence, is estimated to be between 200,000 to 20 million (20,000 genes × 10-1,000 mutations per gene). The variations collected by the major consortia are assured of being in databases, as such deposition is part of the funding conditions, including for the now ubiquitous Genome-Wide Association Studies (GWAS) sphere in the general area of translational medicine. In fact, there is a huge chasm in understanding and funding between the common disease consortia and the necessarily fragmented inherited disease world. The HVP hopes to enhance the funding and the recognition of this latter field.

One target in the field of high-throughput sequencing activity is personal genomics. Currently, this is focused on potential drug dosing and the elucidation of any small risk increases if a particular variation is present. Those receiving their own sequence, however, will need to know the significance of, say, changes in the breast cancer antigen or colon cancer genes. This is where a list of pathogenic mutations pursued by the HVP will be needed. Further, two individuals planning to have children will be able to match their sequence to see if they each have damaging mutation in any one of their genes which might cause recessive disease. If so, pre-implantation/prenatal counselling will be indicated in an analogous way to that for thalassaemia in Greece.

So how is the project 'to collect all mutations in all genes from all countries' progressing? This seems to be a massive task but because the data are needed, it needs to be, and will be, done. It is simply a case of working out how, and spreading the load to perhaps 10,000 people in, say, 100 countries.

There are two critical pilot studies under way which will provide guidance for the whole project.