Decipher the Genome: Understand the Ab Initio Method in Gene Prediction
One of the foundational aspect of genomics to decode the genetic blueprint of species is gene prediction. The problem of accurate interpretation of sequence makes an appearance when the task of identifying genes from a vast stretch of DNA arises. Ab initio is one of the most influential methodology of gene prediction that stands out for this endeavor.
Ab Initio Method:
Ab Initio in gene prediction explains the approach to anticipate gene that solely depends on the DNA sequence which is extracted from genome in the absence of any assistance from experimental data or gene sequences. This method uses statistical approach to recognize gene formation within the genome. “Ab Initio” itself means “from the beginning” which derives from Latin and this method stands true to its meaning.
Functioning of Ab Initio Gene Prediction:
The working of ab initio methods and its models rely on statistical modelling, Hidden Markov Models (HMM) and its training with already known genes.
Statistical Models: The ab initio method analyzes DNA sequences on the basis of its statistical models. The architecture of these models depends on the typical characters of a gene which is the existence of start and stop codons, coding and non-coding sequences also known as exons and introns respectively. These features help to predict the particular required region of a gene.
Hidden Markov Models (HMM): Hidden Markov Model is a powerful technique as it can depict the probabilistic essence of a biological sequence. In gene prediction exons, introns and other regions of a gene can be modeled by HMM which results to foresee a gene structure.
Tutoring with Known Genes: Before launching an ab initio model, it is trained with known sequences of a dataset so it can gain understanding of statistical features of different genomic characters which can be then used to picture genes in the latest sequences.
Ab initio Algorithms and tools:
Many algorithms and programs are designed to facilitate the Ab initio method. Ones that are most widely used are:
GeneMark: For the successful annotation of prokaryotic genome GeneMark is preferred as it differentiates between coding and non-coding zones present in a sequence using Hidden Markov Model (HMM).
GENSCAN: This program is competent for predicting exon-intron structures and multiple genes in a DNA sequence. GENSCAN is favored for the thriving annotation of eukaryotic genome.
GlimmmerHMM: This is also a eukaryotic genome gene prediction tool. It uses HMM to evaluate the positions of genes through a DNA sequence which improves the prediction accuracy.
Supremacy of Ab Initio method:
Ab initio method yields various advantages:
Liberty from exterior data: Ab initio method does not ask for prior information about gene sequences or experimental data for the genome that is being scrutinized which makes it ideal for recently sequenced species with limited experimental data available.
Highly considerate: This method is highly sensitive which makes it to identify genes that are not noticed by evidence based methods, particularly for those genomes with distinctive gene anatomy.
Scalability: Ab initio method can be employed on complete genomes which may qualify it for broad genomic projects.
Restriction and Challenges:
Despite of being influential, the ab initio method is equipped with some limitations. The basic challenge it faces is regarding its possibility for false positives and false negatives which questions its accuracy. It does not provide authentication either the genes which are predicted are functional or expressed. And most importantly it is unfit to predict the eukaryotic genomes which hold complex gene structures accompanied with overlapping genes and alternative splicing.
Future of Gene Prediction:
With the evolution of sequencing technology, there is exponential increase in the genomic data which calls for efficient gene prediction methods. Ab initio method provides evidence based results with hybrid methods which are created when it is merged with comparative genomics or RNA-Seq. It will initiate potentially precise predictions of gene function and regulation with advancements in machine learning and datasets for ab initio model giving a broad genomic landscape.
Conclusion:
Ab initio method of gene prediction allows researchers to locate genes in genomes that are newly sequenced. The discovery of features of DNA sequences that were not known recently has paved the path for incredible researches. As the field of gene prediction continues to make progress, the pledge of decrypting the genome’s unrevealed side becomes ever more accomplishable.
References:
Alioto, T. (2012). Gene Prediction. In Methods in Molecular Biology (Clifton, N.J.). Methods in Molecular Biology (pp. 175–201). doi:10.1007/978-1-61779-582-4_6
Wang, Z., Chen, Y., & Li, Y. (2004). A brief review of computational gene prediction methods. Genomics, Proteomics & Bioinformatics, 2(4), 216–221. doi:10.1016/s1672-0229(04)02028-5
Stanke, M., & Waack, S. (2003). Gene prediction with a hidden Markov model and a new intron submodel. Bioinformatics-Oxford, 19(2), 215-225.
Do, J. H., & Choi, D. K. (2006). Computational approaches to gene prediction. The Journal of Microbiology, 44(2),
Picardi, E., & Pesole, G. (2010). Computational methods for ab initio and comparative gene finding. Data mining techniques for the life sciences, 269-284.
Meyer, I. M., & Durbin, R. (2002). Comparative ab initio prediction of gene structures using pair HMMs. Bioinformatics (Oxford, England), 18(10), 1309–1318. doi:10.1093/bioinformatics/18.10.1309
