We have been using PHRAP to assemble a 20+ Mbp shotgun sequence database
as part of the Giardia lamblia Genome Project (www.mbl.edu/Giardia). We
have additionally started using it to assemble the results of some
directed sequencing on cosmid inserts. For one cosmid of interest, we
get false assemblies when using forcelevel of 1. When we increase the
forcelevel to 3, we get an accurate assembly. We see similar effects
when assembling the 20+ Mbp dataset. The online PHRAP manual leads us
to believe that forcelevel should behave in the opposite manner (text
attached below). Does anyone have any insight on the behaviour of
Andrew G. McArthur, Ph.D.
Marine Biological Laboratory
mcarthur at evol5.mbl.edu
---- from online PHRAP manual ----
(phrap only). Relaxes stringency to varying degree during final contig
merge pass. Allowed values are integers from 0 (most stringent) to 10
(least stringent), inclusive.