Cosmid libraries
A pilot library (called the "Brad Pitt series)
was picked and grdded during the Filarial Genome Project meeting in
1997. This library was
end-sequenced,
and appears to contain
Brugia genomic
DNA. Only 6 96-well plates were
picked.
A second round of cloning was undertaken, and an
18,000 clone cosmid library (insert size 35 kb) was constructed by
Andrea in Steve Williams' lab. Theoretically this yields a 6-fold
coverage of the genome. This library was gridded and has been tested
in several ways.
Stability (D. Guiliano, Edinburgh) A number of cosmid clones were selected at random and grown multiple times (10). DNA was prepared from the separate cultures and analysed by restriction fingerprinting. All the cosmids were stable.
Representativity (D. Guiliano, Edinburgh). A number of different probes from the Brugia nuclear genome were hybridised to the filter and the number of positives identified compared to that expected.
Hha I repeat expect ~10 % found 0 ssu rRNA expect 1% found 3From this we conclude that the library is not representative. Failure of other single-copy probes to hybridise to the filter concurr with this.
Sequencing (Jennifer Wade, NEB and D. Guiliano, Edinburgh). If the library is from Brugia nuclear DNA we would expect its AT content to be ~75%. Sequencing of the ends of 50 cosmids revealed an AT content of~50% and several significant hits to bacterial genes. These bacterial inserts were not E. coli, and do not appear to derive from the endosymbiont (as they do not hybridise to endosymbiont genes). .
We conclude that the gridded cosmid library is not a good substrate for mapping the Brugia nuclear genome. It appears to be ~95% contaminant bacterial DNA.
A second large-scale
cosmid library is under construction in Barton Slatko's lab. (March
1999)