Limits...
Evolutionary Modeling of Genotype-Phenotype Associations, and Application to Primate Coding and Non-coding mtDNA Rate Variation.

O'Connor TD, Mundy NI - Evol. Bioinform. Online (2013)

Bottom Line: We also found a significant association between maximum lifespan and the evolutionary rate of the control region of the mtDNA.In contrast, 24 protein-coding genes from the nuclear genome do not show a consistent pattern of association, which is inconsistent with the generation time hypothesis.These results show the extended method can robustly identify genotype-phenotype associations up to at least four phenotypic categories, and demonstrate the successful application of the method to study factors affecting neutral evolutionary rate in protein-coding and non-coding loci.

View Article: PubMed Central - PubMed

Affiliation: Department of Genome Sciences, University of Washington, Seattle, WA, 98195, USA.

ABSTRACT
Variation in substitution rates across a phylogeny can be indicative of shifts in the evolutionary dynamics of a protein or non-protein coding regions. One way to understand these signals is to seek the phenotypic correlates of rate variation. Here, we extended a previously published likelihood method designed to detect evolutionary associations between genotypic evolutionary rate and phenotype over a phylogeny. In simulation with two discrete categories of phenotype, the method has a low false-positive rate and detects greater than 80% of true-positives with a tree length of three or greater and a three-fold or greater change in substitution rate given the phenotype. In addition, we successfully extend the test from two to four phenotype categories and evaluated its performance. We then applied the method to two major hypotheses for rate variation in the mitochondrial genome of primates-longevity and generation time as well as body mass which is correlated with many aspects of life history-using three categories of phenotype through discretization of continuous values. Similar to previous results for mammals, we find that the majority of mitochondrial protein-coding genes show associations consistent with the longevity and body mass predictions and that the predominant signal of association comes from the third codon position. We also found a significant association between maximum lifespan and the evolutionary rate of the control region of the mtDNA. In contrast, 24 protein-coding genes from the nuclear genome do not show a consistent pattern of association, which is inconsistent with the generation time hypothesis. These results show the extended method can robustly identify genotype-phenotype associations up to at least four phenotypic categories, and demonstrate the successful application of the method to study factors affecting neutral evolutionary rate in protein-coding and non-coding loci.

No MeSH data available.


Results of simulation for binary phenotype. The Z-axis is percentage significant out of 50 tests done for each permutation of tree length (sum of all branch lengths) and W1. Tree length and W1 take on values from 1 to 10 at each integer. All tests have W0 set to 1, a 16 taxa tree, and 50% sites associated.
© Copyright Policy - open-access
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3733722&req=5

f1-ebo-9-2013-301: Results of simulation for binary phenotype. The Z-axis is percentage significant out of 50 tests done for each permutation of tree length (sum of all branch lengths) and W1. Tree length and W1 take on values from 1 to 10 at each integer. All tests have W0 set to 1, a 16 taxa tree, and 50% sites associated.

Mentions: Figure 1 shows our first reevaluation of the method presented in O’Connor and Mundy16 using a revised simulation program. Under this new, more intuitive method of simulation the binary case has a high level of power (>80%) for tree lengths greater than three and a W1 value greater than three. This is important as the results reported earlier required a 10-fold change in molecular rate in order to observe any level of power. The power of the method increases quite dramatically, but similarly to other likelihood methods, discriminating between and weak alternative models can be difficult.67 Here, a doubling of the rate can still be detected when given enough evolutionary divergence, as measured by tree length. For example, with a tree length of 1, 5, and 10 and W1 = 2 we observe 12%, 40%, and 82% significance, respectively.


Evolutionary Modeling of Genotype-Phenotype Associations, and Application to Primate Coding and Non-coding mtDNA Rate Variation.

O'Connor TD, Mundy NI - Evol. Bioinform. Online (2013)

Results of simulation for binary phenotype. The Z-axis is percentage significant out of 50 tests done for each permutation of tree length (sum of all branch lengths) and W1. Tree length and W1 take on values from 1 to 10 at each integer. All tests have W0 set to 1, a 16 taxa tree, and 50% sites associated.
© Copyright Policy - open-access
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3733722&req=5

f1-ebo-9-2013-301: Results of simulation for binary phenotype. The Z-axis is percentage significant out of 50 tests done for each permutation of tree length (sum of all branch lengths) and W1. Tree length and W1 take on values from 1 to 10 at each integer. All tests have W0 set to 1, a 16 taxa tree, and 50% sites associated.
Mentions: Figure 1 shows our first reevaluation of the method presented in O’Connor and Mundy16 using a revised simulation program. Under this new, more intuitive method of simulation the binary case has a high level of power (>80%) for tree lengths greater than three and a W1 value greater than three. This is important as the results reported earlier required a 10-fold change in molecular rate in order to observe any level of power. The power of the method increases quite dramatically, but similarly to other likelihood methods, discriminating between and weak alternative models can be difficult.67 Here, a doubling of the rate can still be detected when given enough evolutionary divergence, as measured by tree length. For example, with a tree length of 1, 5, and 10 and W1 = 2 we observe 12%, 40%, and 82% significance, respectively.

Bottom Line: We also found a significant association between maximum lifespan and the evolutionary rate of the control region of the mtDNA.In contrast, 24 protein-coding genes from the nuclear genome do not show a consistent pattern of association, which is inconsistent with the generation time hypothesis.These results show the extended method can robustly identify genotype-phenotype associations up to at least four phenotypic categories, and demonstrate the successful application of the method to study factors affecting neutral evolutionary rate in protein-coding and non-coding loci.

View Article: PubMed Central - PubMed

Affiliation: Department of Genome Sciences, University of Washington, Seattle, WA, 98195, USA.

ABSTRACT
Variation in substitution rates across a phylogeny can be indicative of shifts in the evolutionary dynamics of a protein or non-protein coding regions. One way to understand these signals is to seek the phenotypic correlates of rate variation. Here, we extended a previously published likelihood method designed to detect evolutionary associations between genotypic evolutionary rate and phenotype over a phylogeny. In simulation with two discrete categories of phenotype, the method has a low false-positive rate and detects greater than 80% of true-positives with a tree length of three or greater and a three-fold or greater change in substitution rate given the phenotype. In addition, we successfully extend the test from two to four phenotype categories and evaluated its performance. We then applied the method to two major hypotheses for rate variation in the mitochondrial genome of primates-longevity and generation time as well as body mass which is correlated with many aspects of life history-using three categories of phenotype through discretization of continuous values. Similar to previous results for mammals, we find that the majority of mitochondrial protein-coding genes show associations consistent with the longevity and body mass predictions and that the predominant signal of association comes from the third codon position. We also found a significant association between maximum lifespan and the evolutionary rate of the control region of the mtDNA. In contrast, 24 protein-coding genes from the nuclear genome do not show a consistent pattern of association, which is inconsistent with the generation time hypothesis. These results show the extended method can robustly identify genotype-phenotype associations up to at least four phenotypic categories, and demonstrate the successful application of the method to study factors affecting neutral evolutionary rate in protein-coding and non-coding loci.

No MeSH data available.