Limits...
Dissecting dynamic genetic variation that controls temporal gene response in yeast.

Brodt A, Botzman M, David E, Gat-Viks I - PLoS Comput. Biol. (2014)

Bottom Line: Here we develop a computational procedure that captures temporal changes in genetic effects, and apply it to analyze transcription during inhibition of the TOR signaling pathway in segregating yeast cells.We found a high-order coordination of gene modules: sets of genes co-associated with the same genetic variant and sharing a common temporal genetic effect pattern.Our analysis suggests that the same mechanism typically leads to both inter-individual variation and the temporal genetic effect pattern in a module.

View Article: PubMed Central - PubMed

Affiliation: Department of Cell Research and Immunology, Tel Aviv University, Tel Aviv, Israel.

ABSTRACT
Inter-individual variation in regulatory circuits controlling gene expression is a powerful source of functional information. The study of associations among genetic variants and gene expression provides important insights about cell circuitry but cannot specify whether and when potential variants dynamically alter their genetic effect during the course of response. Here we develop a computational procedure that captures temporal changes in genetic effects, and apply it to analyze transcription during inhibition of the TOR signaling pathway in segregating yeast cells. We found a high-order coordination of gene modules: sets of genes co-associated with the same genetic variant and sharing a common temporal genetic effect pattern. The temporal genetic effects of some modules represented a single state-transitioning pattern; for example, at 10-30 minutes following stimulation, genetic effects in the phosphate utilization module attained a characteristic transition to a new steady state. In contrast, another module showed an impulse pattern of genetic effects; for example, in the poor nitrogen sources utilization module, a spike up of a genetic effect at 10-20 minutes following stimulation reflected inter-individual variation in the timing (rather than magnitude) of response. Our analysis suggests that the same mechanism typically leads to both inter-individual variation and the temporal genetic effect pattern in a module. Our methodology provides a quantitative genetic approach to studying the molecular mechanisms that shape dynamic changes in transcriptional responses.

Show MeSH

Related in: MedlinePlus

The DyVER algorithm.A methodology for reconstructing genetic associations and their temporal genetic effect patterns from gene expression and genotyping data. (A) A cartoon example of input data, including the expression of a single gene over time for strains s1–s4 (top panel; shown as in Fig. 1A), and a typical genotyping of (homozygous) strains carrying either the  (brown) or  (black) genotype in each genomic position (bottom panel). Correct and incorrect variants (v, u, respectively) are highlighted. (B) Shown are observed effect matrices for each time point from t1 to t4 (red, high-effect size; white, low-effect size). DyVER calculates the observed effects between each pair of strains carrying distinct alleles (strains carrying aa or  in columns and rows, respectively), using a variant u (left) or v (right). (C) Searching for the temporal two-state model that best fits the data. Shown are four cases, for two possible variants u, v, and two possible two-state models. The two states are ‘H’ (light blue) and ‘L’ (white) indicating high and low genetic effect, respectively. DyVER's fit of observed effects (high or low) in two Gaussians and the respective likelihood scores are presented in each case. For each variant, DyVER uses an HMM-based dynamic programming to identify its best-likelihood effect pattern. (D) A Manhattan plot of DyVER scores. Shown are likelihood ratio scores, called DyVER scores (y-axis), quantifying each variant (x-axis) with its selected temporal two-state model (from C). A dashed line indicates the significance threshold, generated using a permutation test.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4256076&req=5

pcbi-1003984-g002: The DyVER algorithm.A methodology for reconstructing genetic associations and their temporal genetic effect patterns from gene expression and genotyping data. (A) A cartoon example of input data, including the expression of a single gene over time for strains s1–s4 (top panel; shown as in Fig. 1A), and a typical genotyping of (homozygous) strains carrying either the (brown) or (black) genotype in each genomic position (bottom panel). Correct and incorrect variants (v, u, respectively) are highlighted. (B) Shown are observed effect matrices for each time point from t1 to t4 (red, high-effect size; white, low-effect size). DyVER calculates the observed effects between each pair of strains carrying distinct alleles (strains carrying aa or in columns and rows, respectively), using a variant u (left) or v (right). (C) Searching for the temporal two-state model that best fits the data. Shown are four cases, for two possible variants u, v, and two possible two-state models. The two states are ‘H’ (light blue) and ‘L’ (white) indicating high and low genetic effect, respectively. DyVER's fit of observed effects (high or low) in two Gaussians and the respective likelihood scores are presented in each case. For each variant, DyVER uses an HMM-based dynamic programming to identify its best-likelihood effect pattern. (D) A Manhattan plot of DyVER scores. Shown are likelihood ratio scores, called DyVER scores (y-axis), quantifying each variant (x-axis) with its selected temporal two-state model (from C). A dashed line indicates the significance threshold, generated using a permutation test.

Mentions: We devised a new method, DyVER, to identify genetic variants that underlie the expression of genes and their particular dynamic effect patterns. DyVER takes as input the measured transcription response of a gene over several consecutive time points following stimulation and across a cohort, as well as a set of potential genetic variants and their genotyping (Fig. 2A). Given a candidate genetic variant with two alternative alleles, DyVER proceeds in three steps (Methods): (1) It first calculates the observed effect of the variant, namely the difference in gene response between strains carrying the two distinct alleles (Fig. 2B). The observed genetic effects are used as data in the subsequent steps. (2) To identify non-linear dynamic shapes of genetic effects, DyVER assumes a ‘digital’ regulatory model that distinguishes two possible states of genetic effects: first, a strong effect of genetic variant on the gene response (denoted the high-effect state); and second, a lower (such as zero) effect, or possibly an opposite effect (denoted the low-effect state). Several previous methods have employed a two-state model, although not in a dynamic or a genetic effect context [28]. Based on a maximum likelihood approach, DyVER seeks a genetic variant and a sequence of states that best describe the dynamic changes in the size of the genetic effect. For example, if a gene is affected mainly by a variant v during a late time interval, DyVER successfully infers the correct effect pattern low→low→high→high for the correct variant v as it attains the highest likelihood score (Fig. 2B and C, right panel). For incorrect variants, the likelihood scores are typically lower (Fig. 2B and C, left panel). DyVER's predicted sequence of states is referred to as the temporal two-state model. Finally, (3) DyVER calculates the statistical significance of association for each genetic variant based on a likelihood ratio score that takes as input the inferred temporal two-state model (Fig. 2D). We refer to this score as the DyVER score. Notably, although DyVER requires synchronous observations in particular time points, it is still possible to apply DyVER on partial observations in each of the time points (Methods).


Dissecting dynamic genetic variation that controls temporal gene response in yeast.

Brodt A, Botzman M, David E, Gat-Viks I - PLoS Comput. Biol. (2014)

The DyVER algorithm.A methodology for reconstructing genetic associations and their temporal genetic effect patterns from gene expression and genotyping data. (A) A cartoon example of input data, including the expression of a single gene over time for strains s1–s4 (top panel; shown as in Fig. 1A), and a typical genotyping of (homozygous) strains carrying either the  (brown) or  (black) genotype in each genomic position (bottom panel). Correct and incorrect variants (v, u, respectively) are highlighted. (B) Shown are observed effect matrices for each time point from t1 to t4 (red, high-effect size; white, low-effect size). DyVER calculates the observed effects between each pair of strains carrying distinct alleles (strains carrying aa or  in columns and rows, respectively), using a variant u (left) or v (right). (C) Searching for the temporal two-state model that best fits the data. Shown are four cases, for two possible variants u, v, and two possible two-state models. The two states are ‘H’ (light blue) and ‘L’ (white) indicating high and low genetic effect, respectively. DyVER's fit of observed effects (high or low) in two Gaussians and the respective likelihood scores are presented in each case. For each variant, DyVER uses an HMM-based dynamic programming to identify its best-likelihood effect pattern. (D) A Manhattan plot of DyVER scores. Shown are likelihood ratio scores, called DyVER scores (y-axis), quantifying each variant (x-axis) with its selected temporal two-state model (from C). A dashed line indicates the significance threshold, generated using a permutation test.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4256076&req=5

pcbi-1003984-g002: The DyVER algorithm.A methodology for reconstructing genetic associations and their temporal genetic effect patterns from gene expression and genotyping data. (A) A cartoon example of input data, including the expression of a single gene over time for strains s1–s4 (top panel; shown as in Fig. 1A), and a typical genotyping of (homozygous) strains carrying either the (brown) or (black) genotype in each genomic position (bottom panel). Correct and incorrect variants (v, u, respectively) are highlighted. (B) Shown are observed effect matrices for each time point from t1 to t4 (red, high-effect size; white, low-effect size). DyVER calculates the observed effects between each pair of strains carrying distinct alleles (strains carrying aa or in columns and rows, respectively), using a variant u (left) or v (right). (C) Searching for the temporal two-state model that best fits the data. Shown are four cases, for two possible variants u, v, and two possible two-state models. The two states are ‘H’ (light blue) and ‘L’ (white) indicating high and low genetic effect, respectively. DyVER's fit of observed effects (high or low) in two Gaussians and the respective likelihood scores are presented in each case. For each variant, DyVER uses an HMM-based dynamic programming to identify its best-likelihood effect pattern. (D) A Manhattan plot of DyVER scores. Shown are likelihood ratio scores, called DyVER scores (y-axis), quantifying each variant (x-axis) with its selected temporal two-state model (from C). A dashed line indicates the significance threshold, generated using a permutation test.
Mentions: We devised a new method, DyVER, to identify genetic variants that underlie the expression of genes and their particular dynamic effect patterns. DyVER takes as input the measured transcription response of a gene over several consecutive time points following stimulation and across a cohort, as well as a set of potential genetic variants and their genotyping (Fig. 2A). Given a candidate genetic variant with two alternative alleles, DyVER proceeds in three steps (Methods): (1) It first calculates the observed effect of the variant, namely the difference in gene response between strains carrying the two distinct alleles (Fig. 2B). The observed genetic effects are used as data in the subsequent steps. (2) To identify non-linear dynamic shapes of genetic effects, DyVER assumes a ‘digital’ regulatory model that distinguishes two possible states of genetic effects: first, a strong effect of genetic variant on the gene response (denoted the high-effect state); and second, a lower (such as zero) effect, or possibly an opposite effect (denoted the low-effect state). Several previous methods have employed a two-state model, although not in a dynamic or a genetic effect context [28]. Based on a maximum likelihood approach, DyVER seeks a genetic variant and a sequence of states that best describe the dynamic changes in the size of the genetic effect. For example, if a gene is affected mainly by a variant v during a late time interval, DyVER successfully infers the correct effect pattern low→low→high→high for the correct variant v as it attains the highest likelihood score (Fig. 2B and C, right panel). For incorrect variants, the likelihood scores are typically lower (Fig. 2B and C, left panel). DyVER's predicted sequence of states is referred to as the temporal two-state model. Finally, (3) DyVER calculates the statistical significance of association for each genetic variant based on a likelihood ratio score that takes as input the inferred temporal two-state model (Fig. 2D). We refer to this score as the DyVER score. Notably, although DyVER requires synchronous observations in particular time points, it is still possible to apply DyVER on partial observations in each of the time points (Methods).

Bottom Line: Here we develop a computational procedure that captures temporal changes in genetic effects, and apply it to analyze transcription during inhibition of the TOR signaling pathway in segregating yeast cells.We found a high-order coordination of gene modules: sets of genes co-associated with the same genetic variant and sharing a common temporal genetic effect pattern.Our analysis suggests that the same mechanism typically leads to both inter-individual variation and the temporal genetic effect pattern in a module.

View Article: PubMed Central - PubMed

Affiliation: Department of Cell Research and Immunology, Tel Aviv University, Tel Aviv, Israel.

ABSTRACT
Inter-individual variation in regulatory circuits controlling gene expression is a powerful source of functional information. The study of associations among genetic variants and gene expression provides important insights about cell circuitry but cannot specify whether and when potential variants dynamically alter their genetic effect during the course of response. Here we develop a computational procedure that captures temporal changes in genetic effects, and apply it to analyze transcription during inhibition of the TOR signaling pathway in segregating yeast cells. We found a high-order coordination of gene modules: sets of genes co-associated with the same genetic variant and sharing a common temporal genetic effect pattern. The temporal genetic effects of some modules represented a single state-transitioning pattern; for example, at 10-30 minutes following stimulation, genetic effects in the phosphate utilization module attained a characteristic transition to a new steady state. In contrast, another module showed an impulse pattern of genetic effects; for example, in the poor nitrogen sources utilization module, a spike up of a genetic effect at 10-20 minutes following stimulation reflected inter-individual variation in the timing (rather than magnitude) of response. Our analysis suggests that the same mechanism typically leads to both inter-individual variation and the temporal genetic effect pattern in a module. Our methodology provides a quantitative genetic approach to studying the molecular mechanisms that shape dynamic changes in transcriptional responses.

Show MeSH
Related in: MedlinePlus