Biomedical Engineering Reference
In-Depth Information
struct strands that we
expect
to have at least a given repeat length (as required by
the below protocol). Recall that we have assumed that each database DNA
strand is also redundantly represented by a number (ranging from up to a few
thousand down to as few as 10) of identical DNA strands. This redundancy aids
us here, since this initialization procedure results in a Biomolecular Database,
where most of the redundant strands are lengthened by at least the given multi-
plying factor.
Figure 5 illustrates a scheme for processing genomic DNA into this data-
base format which might include the following steps: (i) Cleave dsDNA into
manageable pieces. (ii) Append prefixes to both ends of both strands. Heat dena-
ture dsDNA. Anneal to circularizing oligo. (iii) Ligate ssDNA circles. (iv) DNA
polymerase reaction with circular templates to produce linear ssDNA containing
multiple concatenated database entries. Note that the process of converting the
DNA into database format may have unintended effects on the representation of
entries in the database due to uneven amplification. Artificial bias may take the
form of variations in the number of copies present on the average strand (distri-
bution of strand lengths) or differences in the number of strands present for a
given database entry. These protocols need to be optimized to take into account
these possible affects.
Figure 5
. Scheme for processing genomic DNA into database format.
Multiple copies of the database entry are required on a single ssDNA strand,
so that when Boolean variables recorded in the prefixes (A and B in Figure 5)
are queried by primer binding and PCR, information recorded farther out toward
the ends of the strand is not lost by failure to be copied (PCR only amplifies
sequence physically between primer binding sites). The goal is to keep at least a
few copies of the prefix information internally within the database strands so
that information is not lost to subsequent rounds of query.
3.9.5. Logical Query Processing Using Repeated PCR Operations
We assume that the logical query is presented as the logical AND of a list of
K
logical clauses (each clause needs to be satisfied), where each clause consists
