13.12.2012 Views

Theoretical and Experimental DNA Computation (Natural ...

Theoretical and Experimental DNA Computation (Natural ...

Theoretical and Experimental DNA Computation (Natural ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

118 5 Physical Implementations<br />

(Fig. 5.7c). We may then remove all intact str<strong>and</strong>s by separating on length using<br />

gel electrophoresis. However, this is not strictly necessary, <strong>and</strong> leaving the<br />

fragmented str<strong>and</strong>s in solution will not affect the operation of the algorithm.<br />

Union<br />

We may obtain the union of two or more tubes by simply mixing their contents<br />

together, forming a single tube.<br />

Copy<br />

We obtain i “copies” of the set U by splitting U into i tubes of equal volume.<br />

We assume that, since the initial tube contains multiple copies of each<br />

c<strong>and</strong>idate str<strong>and</strong>, each tube will also contain many copies.<br />

Select<br />

We can easily detect remaining homogeneous <strong>DNA</strong> using PCR <strong>and</strong> then sequence<br />

str<strong>and</strong>s to reveal the encoded solution to the given problem. One<br />

problem with this method is that there are often multiple correct solutions<br />

left in the soup which must be sequenced using nested PCR. This technique<br />

is only useful when the final solution is known in advance. Also, the use of<br />

PCR may introduce an unacceptable level of error in the read-out procedure.<br />

A possible solution is to use cloning.<br />

Although the initial tube contains multiple copies of each str<strong>and</strong>, after<br />

many remove operations the volume of material may be depleted below an<br />

acceptable empirical level. This difficulty can be avoided by periodic amplification<br />

by PCR (this may also be performed after copy operations).<br />

5.7 Advantages of Our Implementation<br />

As we have shown, algorithms within our model perform successive “filtering”<br />

operations, keeping good str<strong>and</strong>s (i.e., str<strong>and</strong>s encoding a legal solution<br />

to the given problem) <strong>and</strong> destroying bad str<strong>and</strong>s (i.e., those not doing so).<br />

As long as the operations work correctly, the final set of str<strong>and</strong>s will consist<br />

only of good solutions. However, as we have already stated, errors can take<br />

place. If either good str<strong>and</strong>s are accidentally destroyed or bad str<strong>and</strong>s are<br />

left to survive through to the final set, the algorithm will fail. The main advantage<br />

of our model is that it doesn’t repeatedly use the notoriously error<br />

prone separation by <strong>DNA</strong> hybridization method to extract str<strong>and</strong>s containing<br />

a certain subsequence. Restriction enzymes are guaranteed [31, page 9] 2<br />

2 “New Engl<strong>and</strong> Biolabs provides a color-coded 10X NEBuffer with each restriction<br />

endonuclease to ensure optimal (100%) activity.”

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!