A Review of Soft Computing Techniques for Gene Prediction

Table 1

Summary of soft computing techniques for protein-coding gene prediction.

Soft computing technique used	Organism (datasets used)	Program (URLs wherever available)	Prediction type
Back-propagation NN^1	Human, mouse, arabidopsis, drosophila, rice (GenBank [52])	GRAIL-I [21] http://compbio.ornl.gov/grailexp	Exons
Back-propagation NN	Human, vertebrates (GenBank)	GeneParser [24]; http://beagle.colorado.edu/~eesnyder/GeneParser.html	Exons, introns
Back-propagation NN	Human, mouse (GenBank)	GRAIL-II [22]	Exons
Back-propagation NN	Human, mouse, plant (GenBank)	CODEX [26]	Exons
Back-propagation NN	Vertebrates (GenBank)	GIN [29] http://www.bork.emblheidelberg.de/fmilpetz/GIN/	Exons
Back-propagation NN	S. cerevisiae genome (MIPS [53])	MLFANN (yeast genome) [32]	Open reading frames
Back-propagation ANN	Streptococcus pyogenes M group A Streptococcus strains (GenBank)	SpyMGASLacGenePred [35]	Open reading frames
Self-organizing map NN	E. coli, B. subtilis, H. influenza, Buchnera, B. burgdorferi, M. jannaschii, M. genitalium, H. pylori, A. aeolicus, Synechocystis, Y. pestis, D. radiodurans, R. solanacearum, S. coelicolor, C. jejuni (GenBank)	RescueNet [34]; http://bioinf.nuigalway.ie/RescueNet/	Gene-coding region (prokaryotes)
Multilayer perceptron NN	Microbial Genome (DEG [54] NCBI [55])	EG-MLP (microbial genome) [37]	Genes
GA^2	Human genome (GenBank)	Evolutionary algorithm [46]	Exons, introns
NN + GA	E. coli (PromEC [56], Wisconsin-Madison [57])	MultiNNProm [48]	Promoters
NN + GA	Arabidopsis, E. coli, human, mouse, rat (GenBank and HMR195 [58])	RBFN-combining [49]	Exons

1NN (Neural Network), ²GA (Genetic Algorithms).