TY - GEN

T1 - Exon prediction on DNA-Gen Plasmodium falciparum by using Hidden Markov Model

AU - Agoes, Suhartati

AU - Gunawan, Dadang

AU - Sardy, Sar

PY - 2007/12/1

Y1 - 2007/12/1

N2 - This paper will discuss about the exon prediction on DN A-Gen Plasmodium falciparum by using Hidden Markov Model (HMM). The increment of state number was done randomly up to 100, by using a backward-forward HMM with some transition and emission components on each state. Based on the Gen basic structure on coding sequence (CDS), it is developed four expansion models for increasing the state number. The Viterbi algorithm is used for training process, and Viterbi and Baum-Welch algorithms are used for testing process. The correlation coefficient (CC) is applied as a performance indicator for all expansion models. The result of simulation for the basic structure at state 9, show that the best average CC value is 0.73 for Viterbi algorithm, and is 0.72 for Baum-Welch algorithm. In the expansion models, it is found that the average CC value for second expansion model is 0.78 for both algorithms at stage 100. The average processing time for training is shorter at stage 20 and 30, but almost 12-15 times longer at stage 100, while the average processing time for testing by using Baum-Welch algorithm is twice longer than Viterbi algorithm.

AB - This paper will discuss about the exon prediction on DN A-Gen Plasmodium falciparum by using Hidden Markov Model (HMM). The increment of state number was done randomly up to 100, by using a backward-forward HMM with some transition and emission components on each state. Based on the Gen basic structure on coding sequence (CDS), it is developed four expansion models for increasing the state number. The Viterbi algorithm is used for training process, and Viterbi and Baum-Welch algorithms are used for testing process. The correlation coefficient (CC) is applied as a performance indicator for all expansion models. The result of simulation for the basic structure at state 9, show that the best average CC value is 0.73 for Viterbi algorithm, and is 0.72 for Baum-Welch algorithm. In the expansion models, it is found that the average CC value for second expansion model is 0.78 for both algorithms at stage 100. The average processing time for training is shorter at stage 20 and 30, but almost 12-15 times longer at stage 100, while the average processing time for testing by using Baum-Welch algorithm is twice longer than Viterbi algorithm.

KW - Exon prediction

KW - HMM

KW - Plasmodium falciparum

UR - http://www.scopus.com/inward/record.url?scp=84891455872&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84891455872

SN - 1934272078

SN - 9781934272077

T3 - CITSA 2007 - Int. Conference on Cybernetics and Information Technologies, Systems and Applications and CCCT 2007 - Int. Conference on Computing, Communications and Control Technologies, Proceedings

SP - 66

EP - 69

BT - CITSA 2007 - Int. Conference on Cybernetics and Information Technologies, Systems and Applications and CCCT 2007 - Int. Conference on Computing, Communications and Control Technologies, Proceedings

T2 - 4th International Conference on Cybernetics and Information Technologies, Systems and Applications, CITSA 2007, Jointly with the 5th International Conference on Computing, Communications and Control Technologies, CCCT 2007

Y2 - 12 July 2007 through 15 July 2007

ER -