Current methods of gene prediction, their strengths and weaknesses

Abstract

While the genomes of many organisms have been sequenced over the last few years, transforming such raw sequence data into knowledge remains a hard task. A great number of prediction programs have been developed that try to address one part of this problem, which consists of locating the genes along a genome. This paper reviews the existing approaches to predicting genes in eukaryotic genomes and underlines their intrinsic advantages and limitations. The main mathematical models and computational algorithms adopted are also briefly described and the resulting software classified according to both the method and the type of evidence used. Finally, the several difficulties and pitfalls encountered by the programs are detailed, showing that improvements are needed and that new directions must be considered.

Exploring the Plant Transcriptome through Phylogenetic Profiling

“…Despite the high number of publicly available protein and EST sequences for monocots that are extremely valuable for extrinsic gene prediction approaches (Mathé et al, 2002;Allen et al, 2004), these observations indicate that the current gene annotation in rice still suffers from a number of missed genes. In addition, the high number of unclustered rice genes (approximately 8,600 genes) and putative orphans currently lacking any evidence of expression (approximately 7,000 genes) indicate that further improvement and retraining of gene prediction programs, together with newly developed extrinsic gene prediction methods, seems inevitable for fully exploiting the rice genome sequence (Rouzé et al, 1999;Bennetzen et al, 2004).…”

In silico Gene Characterization and biological annotation of Aspergillus niger CBS 513.88

“…Eukaryotic genes can be predicted by using various algorithms such as ab -initio method, homology-based, EST-based etc [6]. Algorithms like HmmGene, Genscan and Fgenesh use Markov chain probabilities by similarity searching from experimental data of predicted genes or existing genes [7]. Here we predicted the protein coding genes of aspergillus by using various computational software providers of FgenesH Algorithm.…”

MYOP: um arcabouço para predição de genes ab initio"

“…Para resumir a informação deste tipo de alinhamento, a abordagem mais simples é através da utilização da seqüência conservada de cada sinal: ATG para o sítio de início de tradução; TAA, TAG, TGA para sítio de fim de tradução; GT para sítio doador; AG para sítio aceitador. Essas seqüências conservadas podem ser utilizadas para facilitar o reconhecimento de sinais biológicos [39], diminuindo o tempo de processamento do algoritmo de predição. Podemos fazer a busca por um tipo de sinal processando apenas as regiões vizinhas da seqüência conservada ao invés da seqüência de entrada inteira.…”

“…Considerando o alto custo financeiro de cada experimento biológico, fornecer interpretações biológicas para todas as seqüências é praticamente impossível sem o desenvolvimento de algoritmos eficientes de reconhecimento de padrões [39].…”

“…Podemos listar cerca de 25 programas com abordagem probabilística [39] dos quais os melhores acertam com precisão somente as estruturas de 20% dos genes [68]. Além disso, uma comparação recente entre programas contemporâneos mostra que não houve uma mudança significativa na performance nos últimos anos [28].…”