Meiotic linkage maps consisting of highly polymorphic, PCR-based markers spanning the human genome have had a tremendous impact on the positional cloning of human disease genes in recent years. The availability of cDNA sequences representing the majority of human genes promises to have an equally dramatic impact on human genetic research over the next few years. Efficient strategies for using cDNA sequence information to identify human disease genes will incorporate partial human cDNA sequence STSs with meiotic linkage maps. However, despite recent technical advances in both physical and genetic mapping, determining the order and distance between large numbers of DNA markers at the 0.5- to 1-Mb resolution level remains a difficult task.

The Stanford Human Genome Center (SHGC) has generated a set of 83 "whole-genome" radiation hybrids (RHs) that has proven very useful for producing high-resolution maps integrating human cDNAs with meiotically ordered polymorphic markers. Each hybrid retains about 18% of the entire human genome with an average fragment size of 4 Mb. Scoring 6000 random markers on this RH set should result in a map of the entire human genome averaging 500-kb resolution, with around 50% of all markers ordered with odds better than 1000:1. To date, 975 markers on the G??n??thon meiotic linkage map have been placed on the RH set at SHGC.

Given that certain human chromosomal fragments show instability over time and in different hybrid passages, SHGC is carrying out all mapping studies using a single large batch of hybrid DNA prepared and distributed to the scientific community by Research Genetics Peter Good-fellow (University of Cambridge, U.K.) and Jean Weissenbach (G??n??thon) have used a lower dose of X rays to generate a second independent set of RH DNAs, distributed as GENEBRIDGE 4 by Research Genetics. These two reagent sets should allow many different laboratories around the world to integrate data in much the same way that DNA from the common set of CEPH families is used to generate an integrated human meiotic linkage map.

To facilitate integration of cDNA markers and meiotic linkage markers, SHGC has established an automated e-mail server for RH mapping information. Rhserver allows scientists who have scored an STS of interest on the SHGC RH panel to determine which of the G??n??thon markers map near their STS. Submitted typing information is subjected to a "two-point" statistical analysis, and rhserver returns a list of markers that link to the subject marker with a LOD score of 6.0 or higher. The list includes linked markers, LOD score of each link, and distance in centirays between linked markers. For this RH set, one centiray is equal to about 30 kb.

Based on the scoring of 313 STSs derived from cDNA sequences, investigators have found that any random marker has a 50% chance of linking with a LOD of 6 or higher to one of the framework markers; this link will represent a valid map assignment more than 97% of the time. As the number of SHGC-scored markers increases, so will the probability of linkage.

Rhserver will be updated periodically to reflect additional scoring information from SHGC. All scoring information sent to rhserver by other laboratories is confidential and is not examined or retained by SHGC. For scientists who wish to carry out additional analyses not provided by rhserver, raw scoring data for the framework markers is available by ftp from shgc.stanford.edu or from the EBI RHdb below.