gene transinfection directs towards gene functional enhancement using genetic algorithm

IERI Procedia 4 ( 2013 ) 268 – 274

2212-6678 © 2013 The Authors. Published by Elsevier B.V.Selection and peer review under responsibility of Information Engineering Research Institutedoi: 10.1016/j.ieri.2013.11.038

2013 International Conference on Electronic Engineering and Computer Science

Gene Transinfection Directs Towards Gene Functional Enhancement Using Genetic Algorithm

Jayanthi Manicassamya,*, Dhavachelvan. Pb aDepartment of computer Science, Pondicherry University, Pondicherry, India. bDepartment of Computer Science, Pondicherry University, Pondicherry, India.

Abstract

Today, Genetic algorithm plays vital role in solving tough optimization problems with constrains for large search spaces. This works on the principle of survival of the fittest, involving natural evolution process of Darwin theory. GAs don’t opt direct method for resolving tough constrain based optimization problems which has to be carried out by some means, through modifying the existing operations in the phases of classical GA. Numerous operators and its operations subsist for which various design issue confront by most researchers. Among these issues one major issue that is hidden from the researches is, carrying out repairing on genes that are faulty. This paper proposes two bio-inspired gene functionality gene transformation and gene transinfection combined together to solve the problem of obtaining optimal solution. This further enhances the gene functionality through repairing in an effortless means. The proposed work has been tested on 0/1 knapsack problem and its results are compared with classical genetic algorithm which yields to provide better results when compared to that of classical GAs. © 2013 Published by Elsevier B.V. Selection and peer review under responsibility of Information Engineering Research Institute Keywords: Bio-inspired Algorithm; Evolutionary Computation; Gene Repair; Genetic Algorithm; Gene; Optimization; Transinfection; 0/1 knapsack problem.

1. Introduction

Today, evolutionary based natural processing plays an important role in solving large constrain based problems for identifying optimal solution where genetic algorithm falls in that category. Genetic Algorithm

Available online at www.sciencedirect.com

© 2013 The Authors. Published by Elsevier B.V.Selection and peer review under responsibility of Information Engineering Research Institute

ScienceDirect

269 Jayanthi Manicassamy and P. Dhavachelvan / IERI Procedia 4 ( 2013 ) 268 – 274

lays its base of “survival of the fittest” behind it, in the process of working of Darwin’s theory [11-13]. In general constrain based problems are hard to consider universal since, they differ from problem to problem. Various operations and measures have been carried out in genetic algorithm to overcome problems faced while solving constrain based problems for attaining optimal solution which could be attained only to a certain extent[4-7].

Based on the analysis carried out on genetic algorithm in depth at each stage and its intermediate processing levels it has been recognized, that the problem lies in viewing and locating it [1-3, 8-10]. In classical genetic algorithm many measures of solutions have been considered by omitting towards considering the intermediate levels between each stage is one top most reason lie behind for not fully attaining solution to constrain based optimization problems for large search space.

This paper, focuses on providing solution to obtain optimal solution through gene enhancement by tuning the outcome attained during crossover based on occurrence of certain constrains. In section 2 backgrounds of Gene transformation and transinfection functionality with its act in attaining optimal solution using GA has been discussed. In section 3 applicability of 0/1 knapsack problem has been explained with result analysis. In section 4 conclusions with future enhancement has been discussed.

2. Gene Transinfection and Transformation Towards Achieving Optimal Solution

Genes are operational Sub-component of DNA which contains specific set of instructions those codes for a particular protein or for a particular function [14]. These genes dwell in chromosomes. Not all genes are found to be functional some may be repressed from their functional part of producing protein synthesis by various means or by various reasons [15]. Normally gene transformation is an adaption in DNA transfer which depends on the activity to be performed for functional establishment [13,14]. This transformation and transinfection could adapt for DNA repair specially that are damaged by enhancing the frequency of the gene expression. This could also be used to normalize the functionality of the system for which assorted researches are currently being researches are currently being carried out in the field of genetics [18]. This paper lies behind this idea in incorporating transformation and transinfection activity in normalizing genes functionality for acquiring best optimized results.

3.1. Gene Transformation and Transinfection In Genetic Algorithm

The concept of gene transformation and transinfection discussed above has been adopted in classical genetic algorithm for attaining optimal solution. This Gene function has been represented as a processing step carried after crossover for obtaining better results compared to other operations carried out in GA. This operation work at gene level on the offspring’s generated after crossover.

Definition 1. Let “GAn” be a set of chromosomes “Cij” generated for the proposed GA with a group of genes ‘Gij’ acting as a testing environment which can be defined as

n ijGA C Gij (1)

Where the following holds ‘GA’ is ‘n’ is the set of chromosome generated for a specific problem. ‘C’ is the chromosome and ‘G’ is subset of genes that reside in each chromosome.

Definition 2. Let “Cij’ be the set of chromosomes selected with subset of ‘Gij’ Genes for performing crossover operation.. GT is transformation and transinfection action performed on genes ‘Gij’ pertaining to constrain violation. This could be defined as


ijC G Tij G CS (2)

Where the following holds ‘C’ is the chromosome and ‘G’ is subset of genes that reside in each chromosome. ‘GT’ action of transinfection and transformation performed on specific chromosome. ‘CS’ constrain hold for the specific problem.

Fig1. Structure of GA with Transformation and Transinfection

Fig1. represents the structure of the genetic algorithm with gene transformation and transinfection as an enhancement in classical genetic algorithm. Based on the fitness value chromosomes are selected for crossing over for generating offspring’s over which gene transformation takes places to roll back to its parental chromosomal structure to perform transinfection activity is any violation of constrains found.

1.1.1. Working Steps of GA with Gene Transformation and Transinfection

Stage1: Generate Initial population. Stage 2: Perform Reproduction a) Generate set of chromosome with different possible outcomes b) Evaluate fitness function for each chromosome. c) Opt any one of the selection technique and work on the chromosome based on fitness function. Stage 3: Generate Offspring’s through crossover operation Stage 4: Evaluate the offspring’s for constrain violation. For constrain violated due to over or lesser

expression. a) Perform gene transformation on generated both offspring’s to gain its parental chromosomal

structure. b) Perform transinfection operation on the modified offspring’s. Stage 5: Perform Mutation if required which is optional. Stage 6: Opt the best offspring and pass it to the next generation. Stage 6: Goto stage 2.

Initialization

Selection

Crossover

Mutation Termination Check

Gene Transformation &Transinfection


All the stages follow classical genetic algorithm except stage 4 is a newly incorporated operating step in classical genetic algorithm for obtaining optimal solution faster.

2.2 Gene Transformation and Transinfection in 0/1 Knapsack problem

Effective working of classical genetic algorithm with gene transformation and transinfection for generating optimal solution has been proved using 0/1 knapsack problem with constrains in this paper. Let “n” be total object taken for consideration ‘Wi’ and ‘Pi’ be the total weight and profit of the selected items. ‘C’ is the total capacity of the knapsack. The objective is to maximize the profit within the capacity of the knapsack with inclusion or exclusion of the object. Binary form of representation has been made in this paper carrying out the process. 1 represents inclusion of the object and 0 represent exclusion of the object.

For 0/1knapsack problem with 3 objects optimal solutions has to be attained within minimal generations Gene Transformation and Transinfection has been shown below for the generated offspring from the sample chromosomes taken..

Generated offspring's Offspring 1 0 1 1 Offspring 2 0 1 0

Fig 2. Generated Offspring’s

Performing Gene Transformation on dissimilar bit based on constrain violation of maximum profit with overweight and minimum profit with minimal weight. This holds to parental chromosomal structure.

Fig 3. Gene Transformation Yet to Occur

Assume that offspring 2 gives max profit with over capacity and offspring gives minimal profit with minimal weight based on the difference in frequency level analyzed on both functionality of profit and weight. Here gene transformation is carried out on the generated offspring to hold the chromosomal structure of their parents. Further gene transinfection is carried on dissimilar pair of bits using equation 5 to enhance the chromosomal structure to certain extent to attain optimal structure. Fig 3, fig 4 and fig 5 represents gene transformation and transinfection carried out on the enhanced offspring’s. The offspring with best generation is taken for the next generation which carries out for n generations.

Generated gene transformation performed offspring’s

Fig 4. Gene Transformation Occurred Perform Gene Transinfection on dissimilar pair of bits on rate of transinfection chosen.

Offspring 1 0 1 0

Offspring 2 0 1 1

Offspring 1 0 1 0

Offspring 2 0 1 1


Fig 5. Gene Transinfection Rate of transinfection chosen at random on mismatch positions is

(3) Where, R is the rate of repair, is ½, – mismatch positions, – Matched Positions and n is the total

locus positions on each gene.

3. Experimental Analysis

Several times, each experimental execution has been carried out on various data sets. Testing done using, Single-point crossover and two-point crossover with minimal mutation rate. Population size varies from problem to problem. In fig 8 best, worst and average case for generating optimal solution for both classical GA (CGA) and enhanced GA (EGA) with transformation and transinfection has been represented.

3.1. Evaluation Scheme 1: Variable in Knapsack Capacity with Fixed objects

In the first evaluation only the capacity of the knapsack will vary but the object will be fixed. The test is carried out for several data sets and tested several times. Fig 6. represented below depict the performance of CGA and EGA for obtaining optimal solution for fixed objects with different knapsack capacity.

3.2. Evaluation Scheme 2: Fixed Knapsack Capacity with Variable Number of Object

In the second evaluation only the capacity of the knapsack is fixed but the number of object will vary. The test is carried out for several data sets and tested several times. Fig 7. represented below depict the performance of CGA and EGA for obtaining optimal solution for set of objects that vary for fixed knapsack capacity.

Offspring 1 0 1 1

Offspring 2 0 0 1

1

2n

iR V


Fig 8. Generating Optimal Solution Best Case, Worst Case and Average Case

Fig 6. Variable in Knapsack Capacity with Fixed Objects

Fig 7. Fixed Knapsack Capacity with Variable Objects


4. Conclusion and Future Enhancement

This paper discusses about two bio-inspired gene functionality, transformation and transinfection for genetic algorithm. Its efficiency in solving constrain based optimization problem proved to yield better results based on the experiment carried out. This transformation and transinfection functions evenly well for supplementary problems also which has not been discussed in this paper. In future for solving real time biological based problems like defective gene identification effectively could be made possible through transformation and transinfection is to be proved. Currently work is been carried out for solving biological based problem through the proposed process.

Acknowledgements

This work is the part of the Research Project under the Major Project Scheme, UGC, India. Reference No: F. No. 40-258/2011(SR), dated 29 June 2011. The authors would like to express their thanks for the financial support offered by the Sponsored Agency.

References

[1] V. K. Koumousis and C. P. Katsaras, “A saw-tooth genetic algorithm combining the effects of variable population size and reinitialization to enhance performance,” IEEE Transactions on Evolutionary Computation, vol. 10, no. 1, pp. 19–28, 2006. [2] David W. Coit and Alice E. Smith, “Use Of A Genetic Algorithm To Optimize A Combinatorial Reliability Design Problem”, 2000. [3] Jim Smith and T.C.Fogarty, “Self Adaptation of Mutation Rates in a Steady State Genetic Algorithm”, IEEExplore, 1996. [4] Heikki Maaranen, Kaisa M iettinen and Antti Pe nttinen, “On initial populations of a genetic algorithm for continuous optimization problems”, J G lob O ptim., springer, pp. 405–436, 2007. [5] Kalyanmoy Deb, “A Fast and Elitist Multiobjective Genetic Algorithm: NSGA-II”, IEEE Transactions On Evolutionary Computation, Vol. 6, No. 2, 2002. [6] A.S. Anagun and T. Sarac, “Optimization of Performance of Genetic Algorithm for 0-1 Knapsack Problems Using Taguchi Method”, LNCS Springer, 2006. [7] Arild Hoff et. al. “Genetic Algorithms for 0/1 Multidimensional Knapsack Problems”, 2005. [8] Richardson, J. , Palmer, R. , Liepins, G. and Hilliard, M., “Some Guidelines for Genetic Algorithms with Penalty Functions”, Proceedings of the Third International Conference on Genetic Algorithms, pp 191-197. [9] Weingartner, H.M. and Ness, D.N, “Methods for the Solution of Multi-dimensional 0/1 Knapsack Problems”, Operations Research 15, pp 83-103, 1993. [10] Domingo, Cesar Hervaz, “Improving Crossover operator for real-coded genetic algorithms using virtual parents”, Springer Journal, 2007. [11] JC Vidal, “An evolutionary approach for learning the weight of relations in linked data”, 2011. [12] Claudio Mattiussi, Markus Waibel and Dario Floreano, “Measures of Diversity for Populations and Distances Between Individuals with Highly Reorganizable Genomes”, Evolutionary Computation, vol. 12 no. 4, pp. 495-515, 2004. [13] George G., David and Mark, “GeneRepair-A repair operator for genetic algorithm” GECCO 2003 conference, July 2005. [14] Amy FitzGerald and Diarmuid, “Genetic repair for optimization under constraints inspired by Arabidopsis Thaliana”, LNCS Springer, 2008. [15] Jayanthi Manicassamy and P. Dhavachelvan, “Automating Disease Diagnosis in human: A time Series Analysis ”, ICWET 2010 - International Conference and Workshop on Emerging Trends in Technology, pp. 798-800, 2010. [16] Jayanthi Manicassamy, “Metrics based quality assessment for retrieval ability of web-based bioinformatics tools”, 2nd International Conference on Computer Science, Engineering and Applications, ICCSEA, pp. 947-958, 2012. [17] C. Gondro and B.P. Kinghorn, “A simple genetic algorithm for multiple sequence alignment”, Journal of genetics and molecular research evolution and technology, vol 6 no. 4, pp. 964-984, 2007. [18] Jayanthi Manicassamy and P. Dhavachelvan, “Metrics based performance control over text mining tools in bioinformatics”, Proceedings of the International Conference on Advances in Computing, Communication and Control, ICAC3'09, pp. 171-176, 2009.

gene transinfection directs towards gene functional enhancement using genetic algorithm

Documents