Limits...
Inferring microRNA regulation of mRNA with partially ordered samples of paired expression data and exogenous prediction algorithms.

Godsey B, Heiser D, Civin C - PLoS ONE (2012)

Bottom Line: It has been shown that the combination of these two approaches gives more reliable results than either by itself.While a few such algorithms give excellent results, none fully addresses expression data sets with a natural ordering of the samples.If the samples in an experiment can be ordered or partially ordered by their expected similarity to one another, such as for time-series or studies of development processes, stages, or types, (e.g. cell type, disease, growth, aging), there are unique opportunities to infer miR-mRNA interactions that may be specific to the underlying processes, and existing methods do not exploit this.

View Article: PubMed Central - PubMed

Affiliation: Department of Statistics and Probability Theory, Vienna University of Technology, Vienna, Austria. briangodsey@gmail.com

ABSTRACT
MicroRNAs (miRs) are known to play an important role in mRNA regulation, often by binding to complementary sequences in "target" mRNAs. Recently, several methods have been developed by which existing sequence-based target predictions can be combined with miR and mRNA expression data to infer true miR-mRNA targeting relationships. It has been shown that the combination of these two approaches gives more reliable results than either by itself. While a few such algorithms give excellent results, none fully addresses expression data sets with a natural ordering of the samples. If the samples in an experiment can be ordered or partially ordered by their expected similarity to one another, such as for time-series or studies of development processes, stages, or types, (e.g. cell type, disease, growth, aging), there are unique opportunities to infer miR-mRNA interactions that may be specific to the underlying processes, and existing methods do not exploit this. We propose an algorithm which specifically addresses [partially] ordered expression data and takes advantage of sample similarities based on the ordering structure. This is done within a Bayesian framework which specifies posterior distributions and therefore statistical significance for each model parameter and latent variable. We apply our model to a previously published expression data set of paired miR and mRNA arrays in five partially ordered conditions, with biological replicates, related to multiple myeloma, and we show how considering potential orderings can improve the inference of miR-mRNA interactions, as measured by existing knowledge about the involved transcripts.

Show MeSH

Related in: MedlinePlus

The top 10 interactions according to the G–O ordering.In the above diagram, we show the miRs (top row) and genes (bottom row) involved in the 10 most significant targeting interactions based on the G–O ordering from Figure 1. In each case, the inferred interaction is negative, meaning that the miR inhibits the expression of the corresponding gene. A red line from an miR to an mRNA indicates that the interaction was predicted by TargetScan and a blue line indicates that the interaction was predicted by miRanda.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3526609&req=5

pone-0051480-g002: The top 10 interactions according to the G–O ordering.In the above diagram, we show the miRs (top row) and genes (bottom row) involved in the 10 most significant targeting interactions based on the G–O ordering from Figure 1. In each case, the inferred interaction is negative, meaning that the miR inhibits the expression of the corresponding gene. A red line from an miR to an mRNA indicates that the interaction was predicted by TargetScan and a blue line indicates that the interaction was predicted by miRanda.

Mentions: Ultimately, our goal with this analysis is to enable the identification of the most promising candidates for further biological investigation. In Figure 2 we show the top ten interactions inferred by the model using the G–O ordering. The three strongest inferred interactions involve the NR3C1 glucocorticoid receptor, which first appears in the 109th interaction on a ranking of interactions by negative correlation. Myeloma patients with low expression of this receptor respond poorly to standard treatment with dexamethasone and have a poor overall prognosis, making this molecule an intrinsically interesting candidate for further investigation [35]. Two of the miRs inferred as targeting this gene, miR-18a and miR-18b (part of the 5th and 6th ranked interactions by negative correlation), share a seed sequence, and are associated with the miR-1792 cluster–a downstream target of the c-myc oncogene [36]. This cluster is well-known to play a role in cancer development as well as normal lymphoid development, and has recently been associated with tumorgenicity in multiple myeloma [37]. The next strongest inferred interaction involves the gene UBE2D3, (targeted by miR-891b) which is a ubiquitin-conjugating enzyme known to be involved in p53 ubiquitination [38]. The next ranked interaction on our list involves the p53 tumor-suppressor (TP53)–an extremely important gene in most, if not all, cancer types–inferred to be targeted by miR-let-7e. Unlike in many cancers, at diagnosis in multiple myeloma, p53 is rarely seen to be mutated or deleted. As it is not changed at the genomic level, it is therefore quite plausible that p53 may be manipulated at the level of translation by miR in this disease, making this pair an intriguing candidate interaction as well.


Inferring microRNA regulation of mRNA with partially ordered samples of paired expression data and exogenous prediction algorithms.

Godsey B, Heiser D, Civin C - PLoS ONE (2012)

The top 10 interactions according to the G–O ordering.In the above diagram, we show the miRs (top row) and genes (bottom row) involved in the 10 most significant targeting interactions based on the G–O ordering from Figure 1. In each case, the inferred interaction is negative, meaning that the miR inhibits the expression of the corresponding gene. A red line from an miR to an mRNA indicates that the interaction was predicted by TargetScan and a blue line indicates that the interaction was predicted by miRanda.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3526609&req=5

pone-0051480-g002: The top 10 interactions according to the G–O ordering.In the above diagram, we show the miRs (top row) and genes (bottom row) involved in the 10 most significant targeting interactions based on the G–O ordering from Figure 1. In each case, the inferred interaction is negative, meaning that the miR inhibits the expression of the corresponding gene. A red line from an miR to an mRNA indicates that the interaction was predicted by TargetScan and a blue line indicates that the interaction was predicted by miRanda.
Mentions: Ultimately, our goal with this analysis is to enable the identification of the most promising candidates for further biological investigation. In Figure 2 we show the top ten interactions inferred by the model using the G–O ordering. The three strongest inferred interactions involve the NR3C1 glucocorticoid receptor, which first appears in the 109th interaction on a ranking of interactions by negative correlation. Myeloma patients with low expression of this receptor respond poorly to standard treatment with dexamethasone and have a poor overall prognosis, making this molecule an intrinsically interesting candidate for further investigation [35]. Two of the miRs inferred as targeting this gene, miR-18a and miR-18b (part of the 5th and 6th ranked interactions by negative correlation), share a seed sequence, and are associated with the miR-1792 cluster–a downstream target of the c-myc oncogene [36]. This cluster is well-known to play a role in cancer development as well as normal lymphoid development, and has recently been associated with tumorgenicity in multiple myeloma [37]. The next strongest inferred interaction involves the gene UBE2D3, (targeted by miR-891b) which is a ubiquitin-conjugating enzyme known to be involved in p53 ubiquitination [38]. The next ranked interaction on our list involves the p53 tumor-suppressor (TP53)–an extremely important gene in most, if not all, cancer types–inferred to be targeted by miR-let-7e. Unlike in many cancers, at diagnosis in multiple myeloma, p53 is rarely seen to be mutated or deleted. As it is not changed at the genomic level, it is therefore quite plausible that p53 may be manipulated at the level of translation by miR in this disease, making this pair an intriguing candidate interaction as well.

Bottom Line: It has been shown that the combination of these two approaches gives more reliable results than either by itself.While a few such algorithms give excellent results, none fully addresses expression data sets with a natural ordering of the samples.If the samples in an experiment can be ordered or partially ordered by their expected similarity to one another, such as for time-series or studies of development processes, stages, or types, (e.g. cell type, disease, growth, aging), there are unique opportunities to infer miR-mRNA interactions that may be specific to the underlying processes, and existing methods do not exploit this.

View Article: PubMed Central - PubMed

Affiliation: Department of Statistics and Probability Theory, Vienna University of Technology, Vienna, Austria. briangodsey@gmail.com

ABSTRACT
MicroRNAs (miRs) are known to play an important role in mRNA regulation, often by binding to complementary sequences in "target" mRNAs. Recently, several methods have been developed by which existing sequence-based target predictions can be combined with miR and mRNA expression data to infer true miR-mRNA targeting relationships. It has been shown that the combination of these two approaches gives more reliable results than either by itself. While a few such algorithms give excellent results, none fully addresses expression data sets with a natural ordering of the samples. If the samples in an experiment can be ordered or partially ordered by their expected similarity to one another, such as for time-series or studies of development processes, stages, or types, (e.g. cell type, disease, growth, aging), there are unique opportunities to infer miR-mRNA interactions that may be specific to the underlying processes, and existing methods do not exploit this. We propose an algorithm which specifically addresses [partially] ordered expression data and takes advantage of sample similarities based on the ordering structure. This is done within a Bayesian framework which specifies posterior distributions and therefore statistical significance for each model parameter and latent variable. We apply our model to a previously published expression data set of paired miR and mRNA arrays in five partially ordered conditions, with biological replicates, related to multiple myeloma, and we show how considering potential orderings can improve the inference of miR-mRNA interactions, as measured by existing knowledge about the involved transcripts.

Show MeSH
Related in: MedlinePlus