17 Genomic Libraries
271
• chromosome telomeres, since they do not end in convenient restriction
sites, and
• fragments flanked by restriction sites that either cleave too infrequently
to allow cloning in currently available vectors or conversely, cleave too
often, below the minimum size requirement for most vectors.
The cloning gaps can be minimized by careful DNA preparation coupled
with careful vector and host selection.
The size of the library, defined as the number of recombinants produced
prior to its amplification, is another parameter that is useful in defining the
depth of a library. Consider a 40 kb partial Mbol human genomic library in
the cosmid PWE 15. Since the totallength of the human genome is -3.0 x 10 6
kb, end-to-end it would be equivalent to -7.5 x 10 4 clones. Partiallibraries,
however, are comprised of overlapping rather than end-to-end segments.
Thus, one aims for 5- 10 fold redundancy (i.e., 5- 10 genome equivalents in
the number of primary recombinants) to have a 95o/o chance of finding a
particular genomic segment in a partial digest library (Seed, Parker and
Davidson 1982). In the example above, -4.0-8.0 x 10 5 primary recombinants would be necessary. The size of the library can be improved by maximizing the efficiency of packaging and transfection/transformation, and by
increasing the quantity of starting material. If a set of restriction sites
cleaves only 5o/o of the time, increasing the number of molecules digested
and cloned will increase the likelihood of finding a clonethat contains a very
short fragment, as presented in Figure 2.
Primary libraries generally are more representative than amplified ones
since clonesthat have a growth advantage become over-represented when
libraries are amplified. Clones with inserts near either the minimum or
maximum size limits may propagate less efficiently than others. Likewise,
clones that include repetitive sequences or encode genes that are transcribed by the host also can have adetrimental effect on growth. As a consequence, a mimimum of 5x the number of primary recombinants should be
evaluated before concluding that a genomic library lacks a particular genomic clone.
Chimerism
A problern common to alllibraries is the coincidental cloning of more than
one fragment in the sameclone (some YAC libraries have a rate of chimerism as high as 30o/o ). Chimerkclones isolated from complete digest libraries
are easily recognized since their digestion with the enzyme used to prepare
271
• chromosome telomeres, since they do not end in convenient restriction
sites, and
• fragments flanked by restriction sites that either cleave too infrequently
to allow cloning in currently available vectors or conversely, cleave too
often, below the minimum size requirement for most vectors.
The cloning gaps can be minimized by careful DNA preparation coupled
with careful vector and host selection.
The size of the library, defined as the number of recombinants produced
prior to its amplification, is another parameter that is useful in defining the
depth of a library. Consider a 40 kb partial Mbol human genomic library in
the cosmid PWE 15. Since the totallength of the human genome is -3.0 x 10 6
kb, end-to-end it would be equivalent to -7.5 x 10 4 clones. Partiallibraries,
however, are comprised of overlapping rather than end-to-end segments.
Thus, one aims for 5- 10 fold redundancy (i.e., 5- 10 genome equivalents in
the number of primary recombinants) to have a 95o/o chance of finding a
particular genomic segment in a partial digest library (Seed, Parker and
Davidson 1982). In the example above, -4.0-8.0 x 10 5 primary recombinants would be necessary. The size of the library can be improved by maximizing the efficiency of packaging and transfection/transformation, and by
increasing the quantity of starting material. If a set of restriction sites
cleaves only 5o/o of the time, increasing the number of molecules digested
and cloned will increase the likelihood of finding a clonethat contains a very
short fragment, as presented in Figure 2.
Primary libraries generally are more representative than amplified ones
since clonesthat have a growth advantage become over-represented when
libraries are amplified. Clones with inserts near either the minimum or
maximum size limits may propagate less efficiently than others. Likewise,
clones that include repetitive sequences or encode genes that are transcribed by the host also can have adetrimental effect on growth. As a consequence, a mimimum of 5x the number of primary recombinants should be
evaluated before concluding that a genomic library lacks a particular genomic clone.
Chimerism
A problern common to alllibraries is the coincidental cloning of more than
one fragment in the sameclone (some YAC libraries have a rate of chimerism as high as 30o/o ). Chimerkclones isolated from complete digest libraries
are easily recognized since their digestion with the enzyme used to prepare