Abstract
BackgroundAs an abstract mapping of the gene regulations in the cell, gene regulatory network is important to both biological research study and practical applications. The reverse engineering of gene regulatory networks from microarray gene expression data is a challenging research problem in systems biology. With the development of biological technologies, multiple time-course gene expression datasets might be collected for a specific gene network under different circumstances. The inference of a gene regulatory network can be improved by integrating these multiple datasets. It is also known that gene expression data may be contaminated with large errors or outliers, which may affect the inference results.ResultsA novel method, Huber group LASSO, is proposed to infer the same underlying network topology from multiple time-course gene expression datasets as well as to take the robustness to large error or outliers into account. To solve the optimization problem involved in the proposed method, an efficient algorithm which combines the ideas of auxiliary function minimization and block descent is developed. A stability selection method is adapted to our method to find a network topology consisting of edges with scores. The proposed method is applied to both simulation datasets and real experimental datasets. It shows that Huber group LASSO outperforms the group LASSO in terms of both areas under receiver operating characteristic curves and areas under the precision-recall curves.ConclusionsThe convergence analysis of the algorithm theoretically shows that the sequence generated from the algorithm converges to the optimal solution of the problem. The simulation and real data examples demonstrate the effectiveness of the Huber group LASSO in integrating multiple time-course gene expression datasets and improving the resistance to large errors or outliers.
Highlights
As an abstract mapping of the gene regulations in the cell, gene regulatory network is important to both biological research study and practical applications
To study the effectiveness of the proposed method, the Huber group LASSO is applied to inferring Gene regulatory network (GRN) from both simulation datasets and real experimental datasets and the results of Huber group LASSO are compared with those from group LASSO in both area under receiver operating characteristic (AUROC) curve and area under the precision and recall (AUPR) curve
From the ROC plots and areas under the receiver operating characteristic (AUROC), it can be seen that the Huber group LASSO performs significantly better than random guess while the group LASSO method is only a little bit better than random guess
Summary
As an abstract mapping of the gene regulations in the cell, gene regulatory network is important to both biological research study and practical applications. The reverse engineering of gene regulatory networks from microarray gene expression data is a challenging research problem in systems biology. With the development of biological technologies, multiple time-course gene expression datasets might be collected for a specific gene network under different circumstances. The inference of a gene regulatory network can be improved by integrating these multiple datasets. It is known that gene expression data may be contaminated with large errors or outliers, which may affect the inference results. Tremendous amount of microarray data that measure expression levels of genes under specific conditions are obtained from experiments. It is a challenging problem in systems biology to reconstruct or “reverse engineer” GRNs by aiming at retrieving the underlying interaction. The lack of observations and the high dimensionality of the data prohibit the direct application of traditional methods and make the inference task extremely challenging
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
