A physical map showing the relative
positions of sets of clones and/or ordered markers is an important tool
for studying the function and structure of the genome. It also provides
templates for further genomic sequencing, which is the ultimate goal of
most genome projects. There are a number of approaches for assembling such
a map. Among those a method utilizing unique short sequence such as sequence-tag-sites
(STSs) is often used. We proposed a maximum likelihood based method for
recovering the order of STSs/clones as well as estimating the distances
for adjacent STSs. This distance model is shown to be equivalent to the
constant retention model used in SEGMAP (Green & Green, 1991). We have
examined the statistical properties of this method, such as the consistency
of the method, and the condition for the plausibility of recovering the
order of STSs/clones, given the incidences between clones and STSs. The
next stage of this ongoing project is to consider the experimental errors
(false positives/negatives of the hybridization/PCR results) in the likelihood
calculation, which would better meet the practical needs. |
