Table of Contents � Why DNA Computing? � The Structure of DNA DNA Computing � Operations on DNA Molecules � Reading DNA � Example of a Molecular Computer Information Processing with DNA Molecules Christian Jacob Why DNA Computing? Challenges of DNA Computing � From silico to carbon. � Biochemical techniques are not yet From microchips to DNA molecules. sufficiently sophisticated or accurate. � Limits to miniaturization with current � Compare Charles Babbage´s „Analytical Engine“ (1810-1820) computer technologies. � Information processing capabilities of organic molecules ... � replace digital switching primitives, � enable new computing paradigms. Still: Why DNA Computing? Key Features of DNA Computing � Massive parallelism of DNA strands � Further reasons to investigate DNA computing: � high density of information storage � support for standard computation � ease of constructing many copies � better understanding of how nature � Watson-Crick complementarity computes � new data structures (molecules) � feature provided „for free“ � new operations � universal twin shuffle language � cut, paste, adjoin, insert, delete, ... � new computability models.
Table of Contents Operations on DNA Molecules � Why DNA Computing? � Separating and fusing DNA strands � The Structure of DNA � Lengthening of DNA � Operations on DNA Molecules � Shortening DNA � Reading DNA � Cutting DNA � Example of a Molecular Computer � Multiplying DNA Enzymes Separating and Fusing DNA Strands Machinery for Nucleotide Manipulation � Denaturation: separating the single � Enzymes are proteins that catalyze chemical reactions. strands without breaking them � Enzymes speed up chemical reactions � weaker hydrogen than phosphodiester extremely efficiently (speedup: 1012) bonding � heat DNA (85° - 90° C) � Enzymes are very specific. � Nature has created a multitude of � Renaturation: enzymes that are useful in processing DNA. � slowly cooling down � annealing of matching, separated strands Lengthening DNA Shortening DNA � DNA polymerase enzymes � DNA nucleases are enzymes add nucleotides to a DNA that degrade DNA. molecule � DNA exonucleases � Requirements: � cleave (remove) nucleotides one at � single-stranded template a time from the ends of the strands � primer, � bonded to the template � Example: Exonuclease III 3´-nuclease � 3´-hydroxyl end available for degrading in 3´-5´direction extension � Note: Terminal transferase needs no primer.
Shortening DNA Cutting DNA � DNA nucleases are enzymes � DNA nucleases are enzymes that degrade DNA. that degrade DNA. � DNA exonucleases � DNA endonucleases � cleave (remove) nucleotides one at � destroy internal phosphodiester a time from the ends of the strands bonds � Example: S1 � Example: Bal31 cuts only single strands or within removes nucleotides from both single strand sections strands � Restriction endonucleases � much more specific � cut only double strands � at a specific set of sites (EcoRI) PCR Multiplying DNA Step 0: Initialization � Amplification of a „small“ amount of a specific � Start with a solution containing the following ingredients: DNA fragment, lost in a huge amount of other pieces. � the target DNA molecule � „Needle in a haystack“ � primers (synthetic oligonucleotides), complementary to the terminal � Solution: PCR = Polymerase Chain Reaction sections � polymerase, � devised by Karl Mullis in 1985 heat resistant � Nobel Prize � nucleotides � a very efficient molecular Xerox machine PCR PCR Step 1: Denaturation Step 2: Priming � Solution heated close to boiling � The solution is cooled down (to temperature (85° - 90° C) . about 55° C). � Primers anneal to their � Hydrogen bonds between the complementary borders. double strands are separated into single strand molecules.
PCR PCR Efficient Xeroxing: 2n copies after n steps Step 3: Extension � The solution is heated again (to about 72° C). Step 1 � A polymerase will extend the primers, Step 2 using nucleotides available in the solution. Step 3 � Two complete strands of the target DNA molecule are Step 4 produced. Step 5 Measuring the Length of DNA Molecules Table of Contents Gel Electrophoresis � DNA molecules are negatively charged. � Why DNA Computing? � Placed in an electric field, they will move � The Structure of DNA towards the positive electrode. � Operations on DNA Molecules � The negative charge is proportional to the length � Reading DNA of the DNA molecule. � The force needed to move the molecule is � Example of a Molecular Computer proportional to its length. � A gel makes the molecules move at different speeds. � DNA molecules are invisible, and must be marked (ethidium bromide, radioactive) Schematic representation of gel electrophoresis Radioactive marker Sequencing a DNA Molecule � Sequencing: � reading the exact sequence of nucleotides comprising a given DNA molecule � based on � the polymerase action of extending a primed single stranded template Ethidium bromide marker � nucleotide analogues � chemically modified � e.g., replace 3´-hydroxyl group (3´-OH) by 3´- hydrogen atom (3´-H) � dideoxynucleotides: - ddA, ddT, ddC, ddG � Sanger method, dideoxy enzymatic method
Sequencing — Part 1 Sequencing — Part 2 � Objective � 4 tubes are prepared: � Tube A, Tube T, Tube C, Tube G � We want to sequence a single stranded molecule � Each of them contains � . � � molecules � Preparation � primers (= compl ( � ) ) � We extend � at the 3´ end by a short (20 bp) � polymerase sequence � , which will act as the W-C � nucleotides A, T, C, and G. complement for the primer compl ( � ). � Tube A contains a limited amount of ddA. � Usually, the primer is labelled (radioactively, or marked � Tube T contains a limited amount of ddT. fluorescently) � Tube C contains a limited amount of ddC. � This results in a molecule � = 3´- �� . � Tube G contains a limited amount of ddG. Reaction in Tube A Resulting Sequences in Tubes � The polymerase enzyme � Tube A: � Tube C: � TCATGCACTGCG � TCATGCACTGCG extends the primer of � , using � TC � TCA the nucleotides present in � TCATGC � TCATGCA Tube A: � TCATGCAC � TCATGCACTGC ddA, A, T, C, G. � Tube T: � using only A, T, C, G: � TCATGCACTGCG � Tube G: � T � � is extended to the full duplex. � TCATGCACTGCG � TCAT � TCATG � using ddA rather than A: � TCATGCACTG � TCATGCACT � complementing will end at the position of the ddA nucleotide. Final Reading of the Strands Table of Contents � Why DNA Computing? � We read: Gel � T � The Structure of DNA electrophoresis: � TC � TCA � Operations on DNA Molecules � TCAT � TCATG � Reading DNA � TCATGC � Tube A: � Tube C: � TCATGCA TCATGCACTGCG � TCATGCACTGCG � � Example of a Molecular Computer � TCATGCAC � TC TCA � � TCATGCACT � TCATGC TCATGCA � � TCATGCACTG � TCATGCAC � TCATGCACTGC � Tube T: � TCATGCACTGC � TCATGCACTGCG TCATGCACTGCG � � Tube G: T � � TCATGCACTGCG TCAT � � TCATG TCATGCACT � � TCATGCACTG
Adleman´s Experiment Hamiltonian Path Example � In 1994 Leonard M. Adleman showed how to vout solve the Hamilton Path Problem, using DNA 4 1 3 computation. vout 0 � Hamiltonian Path Problem: 6 vin 2 5 � Let G be a directed graph with designated input vin and output vertices, vin and vout. � Adleman´s graph � Simplified graph � Find a (Hamiltonian) path from vin to vout that � The only Hamiltonian � Hamiltonian path: involves every vertex in G exactly once. path for this graph is: � Atlanta � Boston � 0—1—2—3—4—5—6 � Chicago � Detroit Vertex and Edge Encodings Adleman´s Algorithm � Each city (node) vi is encoded by two sub-sequences: � vi = vi´ vi´´ � Input: A directed graph G with n vertices, and designated vertices vin and vout. � Each flight (edge) eik from vi to vk is encoded by: � Step 1: Generate paths in G randomly in large � eik = vi´´ vk´ quantities. � Step 2: Reject all paths that Town DNA Name Complement � do not begin with vin and � do not end in vout. � Step 3: Reject all paths that do not involve exactly n Flight DNA Flight Number vertices. � Step 4: For each of the n vertices v : � reject all paths that do not involve v . � Output: YES, if any path remains; NO, otherwise. „DNA Computer“ DNA Computation Performance Evaluation � The town complements and DNA flight numbers are used for � Information density: computation. � 1015 CDs per cm3 � DNA molecules are put in a hydrous solution. � Massively parallel information processing: � Addition of ligase ensures catalysis of phosphodiester bonds. � 106 ops / sec for PCs � Shaking the test tube makes many DNA strands collide and � 1012 ops / sec for supercomputers interact. � 1020 ops / sec possible for DNA � ~1014 computations are carried out in a single second. � DNA computers would be > 1,000,000 times faster than any � computer today. � The solution strand has to be filtered from the test tube: � Energy efficiency: � 2 * 1019 operations per joule for DNA � GCAG TCGG ACTG GGCT ATGT CCGA � Atlanta � Boston � Chicago � Detroit � 109 operations/joule for silicon-based computers
Recommend
More recommend
