Gene Discovery and Localisation

A major goal of all genome projects is to identify, map and characterise the genes. These objectives are prerequisites for advances in biotechnology, medicine and genetics.
1- Laboratory Approaches
Some well-established methods for mapping genes predated the DNA era, including somatic cell hybridisation and family-based linkage analysis. Somatic cell hybridisation was a forerunner of the radiation hybrid mapping panels that played a significant role in the Human Genome Project for high-resolution gene mapping. Somatic cell mapping panels are derived from human–rodent cell hybrids that are formed by fusing
human cells with rodent cells. The hybrids preferentially lose human chromosomes at random, which allows the establishment of a panel of clones that retain different human chromosomes. This permits the assignment of a DNA probe, or an enzyme detected by its activity, to a particular human chromosome. Both somatic cell hybrid analysis and linkage analysis require knowledge of the gene to be mapped and therefore cannot be used alone for gene discovery. However, both approaches can be used to provide low-resolution localisation for a gene: in the case of the somatic cell hybrids, to a particular chromosome, and for linkage analysis, to a linkage group, which may be associated with a chromosome in some instances. A limitation of linkage analysis is that the gene being mapped and the marker genes must be heterozygous in key parent individuals; this was frequently a stumbling block before the availability of polymorphic DNA markers such as microsatellites and SNPs.
Several gene mapping and identification techniques have been developed for use with cloned DNA. A gene can be localised to a chromosome by fluorescence in situ hybridisation (FISH). FISH is carried out on R-banded spreads of metaphase chromosomes using a labelled genomic clone. The clone is hybridised to its complementary sequence on a chromosome and its position revealed by fluorescence confocal microscopy. By alternating the microscope between the hybridised and banded display, it is possible to assign a gene to a particular chromosome band to a resolution of a few megabases.
A high-resolution development of FISH, sometimes called Fibre FISH, can be used to order cloned genes on DNA fibres prepared from chromatin from interphase nuclei. The fibres are made to extend and using the approach it is possible to hybridise three genomic clones simultaneously enabling their relative positions and order to be determined. 96 This is approach can resolve the order of clones that are only about 2–7 kb apart.
Identifying genes de novo in cloned DNA presents a different type of challenge. Several strategies have been developed with varying degrees of success. In some species, expressed genes are associated with upstream unmethylated GpC dinucleotide-rich sequences. This has allowed the use of the restriction endonuclease HpaII to scan cloned DNA for CpG dinucleotide-rich regions. Because this enzyme only cleaves unmethylated cytosine of the CpG dinucleotide, cleavage at such sites may indicate the presence of an expressed gene. Upon digestion with HpaII, these regions form tiny fragments and are known as HpaII tiny fragments (HTF) or CpG islands. Considerable progress in gene localisation and discovery has been made with a PCR-based method known as EST mapping. An EST or Expressed Sequence Tag is a small PCR product that has been generated from a cDNA sequence, thus reflecting an expressed gene in the cell or tissue from which the mRNA was prepared. Public domain ESTs are available from the dbEST database. Release 053008 of dbEST contains more than 50 million ESTs from a large number of species. EST mapping used with large insert genomic clones leads to discovery and provides information on genome organisation, including gene density and localisation.
 ESTs in many species reflect genes of unknown function. For example, a recent study of the brown planthopper, a serious pest of rice plants, generated a library of more than 37 000 nuclear genome ESTs some of which were unrelated to any gene sequences in the databases. These and others could be used to search for genes of relevance to understanding the biology of the pest species and identifying potential target genes for developing novel insecticides against planthoppers.100,101 A further gene isolation system, which uses adapted vectors termed exon trapping or exon amplification, may be used to identify exon sequences. Exon trapping requires the use of a specialised expression vector that will accept fragments of genomic DNA containing sequences for splicing reactions to take place. Following transfection of a eukaryotic cell line, a transcript is produced that may be detected by using specific primers in an RT-PCR. This indicates the nature of the foreign DNA by virtue of the splicing sequences present.

2- Bioinformatics Approaches
Parallel developments in bioinformatics and biocomputing have been essential for the success of the genome projects. In the closing stages of the Human Genome Project, some institutions were generating up to 175 000 individual DNA sequence reads per day, information which needed to analysed, submitted to databases, annotated and checked for quality. A multiplicity of DNA and protein sequences from different species is submitted daily by laboratories worldwide to databases such as GenBank (USA, DNA sequences), EMBL-BANK (Europe, DNA sequences), DDBJ (Japan, DNA sequences) and SwissProt (Europe, protein sequences). All of these ever-expanding and annotated databases are available to the public and the three major DNA sequence databases share information on a daily basis. These massive primary sequence databases have stimulated the emergence of a large number of specialised
genomic and protein databases dedicated to various subjects such as a particular species group, biological features such as disease markers, protein families, protein domains, etc. The Nucleic Acids Research online Molecular Biology Database Collection is a public repository that lists more than 1000 databases.
Many algorithms such as BLAST have been developed to search databases for matching sequences of nucleotides or amino acids and there are now unparalleled opportunities for characterising genes and for studying groups of related genes in silico. It has been estimated that the majority of new cDNA and EST sequences will show similarity to proteins of known function and that many of the sequences will show
similarity to each other. Further precision for identifying gene families is afforded by automated approaches to access databases such as Pfam to search for protein domains.
Other computational approaches have been devised to predict features of genes in otherwise anonymous genomic sequences. Grail is a gene finder program which predicts exons, genes, promoter regions, polyA tails and other features associated with expressed DNA. ‘Cloning’ genes using database resources and dedicated algorithms is now a realityand  biocomputing has become an applied science which enables investigatorsdealing  with sequences held in data repositories to plot a course
that may seldom require the use of conventional laboratory equipment.
Organising and synthesising the exponentially expanding nucleotide data delivered by the second-generation sequencing projects is perhaps the major challenge for genomics.
In addition to the DNA and protein sequence databases, there has been an accompanying development of catalogue databases. Two examples from human genetics are OMIM (Online Mendelian Inheritance in Man), which catalogues salient details and literature references for more than 18 000 human genetic traits (the majority of which a reknown at  the sequence or molecular basis levels), and the HumanGenome Nomenclature Database (HGNC), which assigns names to all known human genes and curates a searchable database. There are man yothers serving various research communities and biological systems.