POA Simulation 2004.5.25 MEC Seminar 임희웅
Contents Introduction Algorithm and Workflow Architecture and Environment Result Discussion Reference
Introduction Lee et al. DNA10, 2004 Inside of POA Prediction of experimental result without real experiment 23 city TSP Information for experimental setup
Algorithm and Workflow Simplified version of Maheshri et al.(2003) Random selection and collision model Selection with a probability proportional to the number of corresponding DNA. (roulette-wheel selection) Annealing event probability by Gibbs free energy and NN model No gapped annealing or bulge Consider only match sections Sum of all the free energy in match sections Boltzmann-weighted probability for each annealing event
Workflow
Architecture and Environment
Result (1) Cycle 10 Cycle 20 Cycle 30 Cycle 6 Cycle 5 Cycle 1 Cycle2
Result (2) 100 200 300
Discussion Fidelity of POA? Annealing temperature Validity of NN model Low extension efficiency in later cycle Unextendable hybridization Annealing temperature Annealing temperature gradient? (low high) Validity of NN model
Reference J. Santalucia, Jr. A unified view of polymer, dumbbell, and oligonucleotide DNA nearest-neighbor thermodynamics, PNAS, 1998 N. Maheshri, et al. Computational and experimental analysis of DNA shuffling, PNAS, 2003, and its supplement J. Y. Lee, et al Efficient initial pool generation for weighted graph problems using parallel overlap assembly, DNA10, 2004