Limits...
Highly sensitive inference of time-delayed gene regulation by network deconvolution.

Chen H, Mundra PA, Zhao LN, Lin F, Zheng J - BMC Syst Biol (2014)

Bottom Line: Experiments on real-life gene expression datasets show that our method achieves overall better performance than existing methods for inferring time-delayed GRNs.The effectiveness of our method has been shown by the experiments on three real-life gene expression datasets of yeast.Compared with other existing methods which were designed for learning time-delayed GRN, our method has significantly higher sensitivity without much reduction of specificity.

View Article: PubMed Central - HTML - PubMed

ABSTRACT

Background: Gene regulatory network (GRN) is a fundamental topic in systems biology. The dynamics of GRN can shed light on the cellular processes, which facilitates the understanding of the mechanisms of diseases when the processes are dysregulated. Accurate reconstruction of GRN could also provide guidelines for experimental biologists. Therefore, inferring gene regulatory networks from high-throughput gene expression data is a central problem in systems biology. However, due to the inherent complexity of gene regulation, noise in measuring the data and the short length of time-series data, it is very challenging to reconstruct accurate GRNs. On the other hand, a better understanding into gene regulation could help to improve the performance of GRN inference. Time delay is one of the most important characteristics of gene regulation. By incorporating the information of time delays, we can achieve more accurate inference of GRN.

Results: In this paper, we propose a method to infer time-delayed gene regulation based on cross-correlation and network deconvolution (ND). First, we employ cross-correlation to obtain the probable time delays for the interactions between each target gene and its potential regulators. Then based on the inferred delays, the technique of ND is applied to identify direct interactions between the target gene and its regulators. Experiments on real-life gene expression datasets show that our method achieves overall better performance than existing methods for inferring time-delayed GRNs.

Conclusion: By taking into account the time delays among gene interactions, our method is able to infer GRN more accurately. The effectiveness of our method has been shown by the experiments on three real-life gene expression datasets of yeast. Compared with other existing methods which were designed for learning time-delayed GRN, our method has significantly higher sensitivity without much reduction of specificity.

Show MeSH
Benchmark network of yeast9. This benchmark network is from [12]. All edges were detected by biological experiments.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4290726&req=5

Figure 1: Benchmark network of yeast9. This benchmark network is from [12]. All edges were detected by biological experiments.

Mentions: We evaluated the performance of our method on two real-life benchmark networks and compared with other related methods. One benchmark network (see Figure 1) is experimentally identified in Saccharomyces cerevisiae (yeast) cell cycle [12]. This network consists of nine genes (ACE2, CLN3, FKH2, MBP1, MCM1, NDD1, SWI4, SWI5 and SWI6). The real expression data (denoted as yeast9 ) of genes in this network were taken from Spellman [13], which consists of the transcription expression data of yeast cell cycle. We extracted the time-series data from cdc-15 cell cycle arrest which contains 24 equally distributed time points. The other benchmark network (see Figure 2) is a five-gene network in yeast, from the experiment of in vivo reverse-engineering and modeling assessment (IRMA) [14]. This network was carefully constructed to include the interactions among five genes (CBF1, GAL4, SWI5, GAL80 and ASH1) in Saccharomyces cerevisiae and made sure that the influence from endogenous genes is negligible. Two datasets of gene expression were measured for this network. One dataset was obtained when the cell culture was shifted from glucose to galactose. This dataset was named "switch-on" because the network would be triggered by galactose. The other dataset was named "switchoff" since it is obtained by shifting the cell culture from galactose to glucose. The "switch-on" dataset (denoted as yeast5on) consists of 16 equally distributed time points, and the "switch-off" dataset (denoted as yeast5off ) contains 21 equally distributed time points.


Highly sensitive inference of time-delayed gene regulation by network deconvolution.

Chen H, Mundra PA, Zhao LN, Lin F, Zheng J - BMC Syst Biol (2014)

Benchmark network of yeast9. This benchmark network is from [12]. All edges were detected by biological experiments.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4290726&req=5

Figure 1: Benchmark network of yeast9. This benchmark network is from [12]. All edges were detected by biological experiments.
Mentions: We evaluated the performance of our method on two real-life benchmark networks and compared with other related methods. One benchmark network (see Figure 1) is experimentally identified in Saccharomyces cerevisiae (yeast) cell cycle [12]. This network consists of nine genes (ACE2, CLN3, FKH2, MBP1, MCM1, NDD1, SWI4, SWI5 and SWI6). The real expression data (denoted as yeast9 ) of genes in this network were taken from Spellman [13], which consists of the transcription expression data of yeast cell cycle. We extracted the time-series data from cdc-15 cell cycle arrest which contains 24 equally distributed time points. The other benchmark network (see Figure 2) is a five-gene network in yeast, from the experiment of in vivo reverse-engineering and modeling assessment (IRMA) [14]. This network was carefully constructed to include the interactions among five genes (CBF1, GAL4, SWI5, GAL80 and ASH1) in Saccharomyces cerevisiae and made sure that the influence from endogenous genes is negligible. Two datasets of gene expression were measured for this network. One dataset was obtained when the cell culture was shifted from glucose to galactose. This dataset was named "switch-on" because the network would be triggered by galactose. The other dataset was named "switchoff" since it is obtained by shifting the cell culture from galactose to glucose. The "switch-on" dataset (denoted as yeast5on) consists of 16 equally distributed time points, and the "switch-off" dataset (denoted as yeast5off ) contains 21 equally distributed time points.

Bottom Line: Experiments on real-life gene expression datasets show that our method achieves overall better performance than existing methods for inferring time-delayed GRNs.The effectiveness of our method has been shown by the experiments on three real-life gene expression datasets of yeast.Compared with other existing methods which were designed for learning time-delayed GRN, our method has significantly higher sensitivity without much reduction of specificity.

View Article: PubMed Central - HTML - PubMed

ABSTRACT

Background: Gene regulatory network (GRN) is a fundamental topic in systems biology. The dynamics of GRN can shed light on the cellular processes, which facilitates the understanding of the mechanisms of diseases when the processes are dysregulated. Accurate reconstruction of GRN could also provide guidelines for experimental biologists. Therefore, inferring gene regulatory networks from high-throughput gene expression data is a central problem in systems biology. However, due to the inherent complexity of gene regulation, noise in measuring the data and the short length of time-series data, it is very challenging to reconstruct accurate GRNs. On the other hand, a better understanding into gene regulation could help to improve the performance of GRN inference. Time delay is one of the most important characteristics of gene regulation. By incorporating the information of time delays, we can achieve more accurate inference of GRN.

Results: In this paper, we propose a method to infer time-delayed gene regulation based on cross-correlation and network deconvolution (ND). First, we employ cross-correlation to obtain the probable time delays for the interactions between each target gene and its potential regulators. Then based on the inferred delays, the technique of ND is applied to identify direct interactions between the target gene and its regulators. Experiments on real-life gene expression datasets show that our method achieves overall better performance than existing methods for inferring time-delayed GRNs.

Conclusion: By taking into account the time delays among gene interactions, our method is able to infer GRN more accurately. The effectiveness of our method has been shown by the experiments on three real-life gene expression datasets of yeast. Compared with other existing methods which were designed for learning time-delayed GRN, our method has significantly higher sensitivity without much reduction of specificity.

Show MeSH