Table of Contents
Potential gene candidates prediction #
Guilt by association #
Guilt by association (GBA) is a principle/rule to predict the Biological functions to unknown genes from available data sets and it's extensively used for Gene expression and Co-expression analyses by biological communities.
Mostly this rule depends on the relationship between two continuous variables (e.g. gene expression levels). The dependence was mostly grouped in to two, i. Positive correlation (when gene B tends to be high when gene A is high) and ii. Negative correlation (when gene B tends to be high when gene A is low).
Indeed, GBA is most essential rule for the biologist to predict the association between the Genotype and Phenotype. From the past experiences, the projects related to genome guided and de novo sequencing projects (Figure 1A), the major problem in the Annotation and selecting the candidate gene for specific objective, because the interested candidates were hidden in the unknown group. Here the partial gene/transcripts were annotation with different level (GO: Gene Ontology, KEGG: Kyoto Encyclopedia of Genes and Genomes) (Figure 1B) with the expression values. With those expression values, the datasets were constructed as a Weighted matrix and converted to networks (Figure 1C). Finally based on the network cluster and cutoff values, the functions were transferred to unknown genes (Figure 1D).
Recent success stories for plants #
Iridoid synthase #
Recently, the co-expression analyses were applied to find the novel gene for the iridoid biosynthesis from Catharanthus roseus from Transcriptome data by applying forward and reverse ranking from calculated Pearson correlation coefficient (PCC) with specific gene (G10H) (Geu-Flores et al., 2012). Another research group (Kazuki Saito from RIKEN, Japan (http://www.csrs.riken.jp/en/labs/mrg/index.html)) were applied the GBA to integrate the metabolomics and transcriptomics and identified 54 novel genes for secondary metabolites in plants.
Other Applications #
GBA also studied for the transcriptome section from Heterogeneous collections of transcriptomics data with respect to GO, KEGG and MIPS functional categories (Bhat et al., 2012), Multifunctional genes were assessed by GBA (Gillis and Pavlidis, 2011) and for the specific gene association with diseases (Howell, 2014) and based on the GBA the computational tools are developed, which are reviewed by (Moreau and Tranchevent, 2012)
Bhat, P., Yang, H., Bogre, L., Devoto, A. and Paccanaro, A., 2012. Computational selection of transcriptomics experiments improves Guilt-by-Association analyses. PLoS One 7, e39681.
Geu-Flores, F., Sherden, N.H., Courdavault, V., Burlat, V., Glenn, W.S., Wu, C., Nims, E., Cui, Y. and O'Connor, S.E., 2012. An alternative route to cyclic terpenes by reductive cyclization in iridoid biosynthesis. Nature 492, 138-42.
Gillis, J. and Pavlidis, P., 2011. The impact of multifunctional genes on "guilt by association" analysis. PLoS One 6, e17258.
Howell, W.M., 2014. HLA and disease: guilt by association. Int J Immunogenet 41, 1-12.
Moreau, Y. and Tranchevent, L.C., 2012. Computational tools for prioritizing candidate genes: boosting disease gene discovery. Nat Rev Genet 13, 523-36.