Linkage has a long history, making it challenging to predict its future. Meiotic recombination of SNPs in small candidate regions holds promise for identifying recombinogenic sequences, but extending this to a whole chromosome remains a significant technical challenge. Such advancements could improve linkage maps precision but would not enhance gene localization without meiotic expression.
Segregation and linkage analysis, with credible adjustments for ascertainment bias, is currently limited to major loci and single markers. However, extending this approach to multiple markers could prove more beneficial than continuously refining weak parametric models. Association mapping, a relatively young technique, has significant potential for expansion. Despite challenges in composite likelihood, Monte Carlo simulations provide greater power for genome scans by linkage than coalescence models.
Genome scans using allelic association offer higher resolution than linkage but demand extensive automation, making them less accessible to most researchers. The competition between microsatellite-based genome scanning and SNP-dominated association mapping continues. Two-dimensional color graphics depicting haplotypes are being evaluated for their effectiveness in association mapping compared to traditional additive distance-based maps. These graphics are highly sensitive to sample size and SNP density, raising questions about their long-term viability.
HapMap, aiming to provide SNP data of broad utility, has yet to determine whether to present its findings as maps. Meanwhile, the study of map determinants is ongoing. Unlike linkage maps, which appear consistent across populations, LD maps are influenced by demographic factors, particularly the time since the last major population bottleneck. Local and cosmopolitan maps vary in reliability, depending on resolution and sample pooling methods.
The construction of LD maps for association mapping of disease susceptibility is generating excitement similar to that of linkage mapping a century ago. In the final stages of association mapping, functional analysis of causal SNPs will become crucial. Haplotype-based analyses may lose relevance as functional tests take precedence in distinguishing association from causation. Eventually, the integration of linkage, LD, and expression analyses will define the future trajectory of genetic mapping.
