The Hindu Business Line : IT is in the DNA too

You were introduced, earlier, to different types of — Autonomic, Affective and Quantum — computing. Here, in a two-part series, we see how the DNA can also become a computing tool.

Dr S Chellaiah

“Lunch is ready. It is the third time I am calling you. Chellaiah, are you there?”

That was my mother. It was a special family lunch. I was “drowned” in the World Wide Web and hence did not pay attention to her.

“How many times should I call you? When you get on the Net, you get submerged”.

“Sorry mom. But it is all in the genes. You know that DNA stuff that made me inherit some qualities from you.”

“What are you saying?”

“Like when cooking, you have 200 per cent concentration. You are focused. I am also …”

“You, genes, DNA and computers….. It is discourteous to make others wait. ”, was her interrupted comment.

My mom, who has learnt from experience far more than I ever did from my long college education, was hitting the point hard.

But, did she say something like DNA computer? Wow! A computer using DNA?

Yes, there sure is a DNA computer. These are computers that use DNA for input, storage and output. They don’t require electricity.

The genesis of computers lies in the ‘Turing Machine’ conceived by the British Mathematician Alan Turing back in 1936. His conceptual machine existed on paper and did not specify any materials to build it. It would read a symbol, change it, then read the next one and continue based on certain rules and would stop when all the rules have been applied.

In living organisms, instructions for their operations are stored in their genomes, in the form of a sequence of nucleic acids. Consider the operations in cells that process DNA. It involves recognition (identification of nucleic acid), cleavage (cutting of DNA at selected points) and ligation (joining of two molecules) and movement along a polymer chain. After scientists understood the biological molecules, the similarities between Turing’s machine and the biological organisms became apparent. This formed the crux of the rationale to consider biological molecules as computers.

Charles Bennett of IBM first proposed the hypothetical molecular Turing machine back in 1982. However, it was only in 1994 that Leonard M. Adleman of the University of Southern California showed that biological molecules can be used to solve traditional problems. (Incidentally, Adleman was one of the trio who invented the RSA encryption algorithm).

DNA – THE WONDER MACROMOLECULE

A nucleotide consists of a sugar (ribose or deoxyribose), a nitrogenous base (an organic compound consisting of carbon and nitrogen) and a phosphate group. There are five common nitrogenous bases called adenine, guanine, cytosine, thymine and uracil. A chain of nucleotide molecules linked together is called a polynucleotide. For most of the living organisms, these chains store information and act as distribution and transmission agents. Genetic information is stored in the form of a sequence of nucleotides.

DNA, an abbreviation for deoxyribonucleic acid, is a large macromolecule consisting of deoxyribose sugar and four nitrogenous bases (adenine, guanine, cytosine, and thymine). Two chains of linked polynucleotides twist around one another in a double helix. A nitrogenous base from one strand pairs with a base on the opposite strand. Adenine base always links with a thymine base and a guanine base links with a cytosine base only. This is denoted as A: T and G: C. They resemble the rungs of a ladder and are called the base pairs.

HOW DO DNA COMPUTERS WORK?

Presently, the DNA computer is in a nascent stage. The idea that DNA can be used for computation has been proved. But there are no DNA computers in commercial use. Hence, what can be described is the concept and experiments that have proven their viability and revealed their awesome potential.

In a DNA computer, DNA is the “software” and the enzymes are the “hardware”. The solution is not displayed on any monitor (screen) but shown as a DNA strand which has to be interpreted by a human being. There are no moving parts and no electricity. The base pairs A, T, C and G constitute the base-four number system in DNA computing. DNA strands such as ATCGGT are used to denote symbols and a state of either “0” or “1”. Boolean operations such as ‘And’ can be done by separating DNA strands and ‘Or’ ” can be done by mixing two DNA solutions together.

By a variety of operations such as separation, merging, extracting a pattern, ligation, amplification with PCR (Polymerase Chain Reaction), and slicing with enzymes, the input DNA strands can be “operated” upon to represent different “states of the variables” or “intermediate values”. For instance, an enzyme will look for select spots on the DNA strand, check for specific information in those spots and if it matches, the enzyme may add “1” to one or both ends of the strand (Here adding 1 denotes the attaching of a nucleotide that represents 1 to the end(s) of the strand). Thus, enzymes are used to do logical and mathematical manipulations.

Further understanding can be obtained by reviewing the pioneering experiment of Adleman.

ADLEMAN’s EXPERIMENT

With a view to capitalise on the molecule’s potential, Adleman attempted to solve the Hamilton Path Problem, also called the Travelling Salesman Problem (refer note 1).

He considered seven cities. Each city was represented by a single stranded DNA molecule containing 20 nucleotides. (These molecules were made using a ‘DNA synthesiser’). Each path that connected two cities was represented by a DNA molecule containing 20 nucleotides composed of the last ten nucleotides of the DNA denoted by the city of departure and the first ten nucleotides of the DNA denoted by the city of arrival. (The DNA strands were joined together by an enzyme called ‘ligase’).

All the paths that connected every city with every other city were represented by DNA strands and were kept in a test tube. Out of these, those strands that have city 1 at one end and city 7 at the other end were selected using Polymerase Chain Reaction (PCR) and kept in another test tube.

This selection then contained double stranded DNA of various lengths, but all had the correct city of departure and the correct city of destination. The various lengths meant that the paths covered only a few of the seven cities. From this group, only those that were seven-city long were selected using a technique called ‘Gel Electrophoresis’. The resulting group of DNA strands was filtered city by city, one city at a time. This yielded a group of DNA strands that encoded each city only once.

A technique called ‘Affinity Purification’ was used to pick a DNA with a specific sequence from the group of DNA strands. Finally, the optimal path represented by a DNA was found using a method called ‘Graduated PCR’. Though the calculations were completed in one second, it took Adleman one week to “dig out the answer from the DNA soup”. (Also see note 2).

Next week: Comparison of traditional and DNA computers and applications of DNA computers.

Note 1: In the travelling salesman problem, a salesperson has to visit a fixed number of cities that are interconnected by limited roads without traversing through any city more than once. The optimal route must be determined. If there are 50 cities, then there will be 250 combinations that need to be considered to arrive at the solution. Even for today’s powerful computers, this will take a long time. This problem belongs to a class called hard NP-complete problems which cannot be solved easily.

With an increase in the number of variables, the time needed to solve these problems increases exponentially. ‘Knight Problem’ (a chess puzzle) that determines the number of knights and the locations where they can be placed so that they don’t attack each other is an example of a hard NP-complete problem.

Note 2: In March 2002, Adleman and his team solved 20 variable NP-complete problem using DNA computers. This is a significant increase over the seven variable problems solved in 1994.

(The author is Consultant, Satyam Computer Services Ltd, and can be reached at chellaiah_s@satyam.com )

More Stories on : Hardware | Software | Science & Technology

Article E-Mail :: Comment :: Syndication :: Printer Friendly Page

Stories in this Section
IT is in the DNA too

It's Telepresence

For that edge in the exam

Going the M-ad way

Net connection

India plus China in ICT

How ‘predictive’ can you get?

Towards value

Quiz

Connected we survive