Figure 4: The expected number of contigs after n steps using a directed strategy. It's Fig 6. in Nelson and Speed (1994).
Instead of using the random approaches such as bottom-up fingerprinting and STS-content mapping, Mizukami et al and Palazzolo et al ([12], [14]) proposed a directed approach to construct the maps. Roughly speaking, directed approaches proceed sequentially by repeatedly
with mean L for a genome of length
G. The expected number and size of contigs and the expected proportion of the
genome left uncovered after step n can be computed ``exactly'' for
the case of constant clone length, and approximately by the similar
thinning arguments. For the formulas and details, see
[10]. Interestingly, the progression of the project
(the number of contigs) behaves quite differently from the random
strategies. See Figure 4(Fig 6. in Nelson and Speed
[10]). Unlike the long right-hand
tail in the bottom-up random
schemes, a project using a directed strategy has a slower progress in
the beginning, but closes the gaps much faster in later stages.