# XGRN: Reconstruction of Biological Networks Based on Boosted Trees Regression

^{1}

^{2}

## Abstract

**:**

## 1. Introduction

## 2. Materials and Methods

#### 2.1. Extreme Gradient Boosting (XGBoost)

#### 2.2. GRN Reconstruction

^{2}. A low error indicates that the samples in testing show similar patterns as in training. Intuitively, we attempt to learn the function “B is regulated by A”, thus as A, we will use profiles of TFs and as B the target genes. Since a TF usually has more than one known target, we train several models (one for each interaction), and hence we obtain multiple prediction values for a candidate target. We combine them by keeping the minimum error (for MSE or MAE) or the maximum value for R

^{2}as the final prediction score.

#### 2.3. Data

#### 2.4. Evaluation

## 3. Results

^{2}metric. Thus this interaction was predicted as true. In Figure 2b, the same procedure was repeated, but in testing, the gene was not a target of this TF. In this case, the output of regression had a low R

^{2}value. Thus this potential interaction was correctly considered false.

#### 3.1. Parameter Selection

^{2}, mean squared error (MSE) and mean absolute error (MAE) as the output of our method when comparing the predicted with the actual expression (Table 2). R

^{2}led to superior results, although the difference was small. This can be explained by the fact that R

^{2}has an upper limit of 1, making easier comparisons among different models, while the magnitude of the other two metrics is affected by the expression levels.

#### 3.2. GRN Inference Performance

#### 3.3. Effect of Prior Knowledge Percentage

## 4. Discussion

## Funding

## Data Availability Statement

## Conflicts of Interest

## References

- Ideker, T.; Krogan, N.J. Differential network biology. Mol. Syst. Biol.
**2012**, 8. [Google Scholar] [CrossRef] [PubMed] - Vidal, M.; Cusick, M.E.; Barabási, A.L. Interactome networks and human disease. Cell
**2011**, 144, 986–998. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Lee, D.S.; Park, J.; Kay, K.A.; Christakis, N.A.; Oltvai, Z.N.; Barabasi, A.-L. The implications of human metabolic network topology for disease comorbidity. Proc. Natl. Acad. Sci. USA
**2008**, 105, 9880–9885. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Csermely, P.; Korcsmáros, T.; Kiss, H.J.M.; London, G.; Nussinov, R. Structure and dynamics of molecular networks: A novel paradigm of drug discovery: A comprehensive review. Pharmacol. Ther.
**2013**, 138, 333–408. [Google Scholar] [CrossRef] [Green Version] - Liu, Q.; Muglia, L.J.; Huang, L.F. Network as a biomarker: A novel network-based sparse bayesian machine for pathway-driven drug response prediction. Genes (Basel)
**2019**, 10, 602. [Google Scholar] [CrossRef] [Green Version] - Network Medicine; Loscalzo, J.; Barabási, A.-L.; Silverman, E.K. (Eds.) Harvard University Press: London, UK, 2017; ISBN 9780674545533. [Google Scholar] [CrossRef]
- Dimitrakopoulou, K.; Dimitrakopoulos, G.N.; Sgarbas, K.N.; Bezerianos, A. Tamoxifen integromics and personalized medicine: Dynamic modular transformations underpinning response to tamoxifen in breast cancer treatment. OMICS
**2014**, 18, 15–33. [Google Scholar] [CrossRef] - Dimitrakopoulou, K.; Dimitrakopoulos, G.N.; Wilk, E.; Tsimpouris, C.; Sgarbas, K.N.; Schughart, K.; Bezerianos, A. Influenza a immunomics and public health omics: The dynamic pathway interplay in host response to H1N1 infection. Omi. A J. Integr. Biol.
**2014**, 18, 167–183. [Google Scholar] [CrossRef] - Langfelder, P.; Horvath, S. WGCNA: An R package for weighted correlation network analysis. BMC Bioinform.
**2008**, 9, 559. [Google Scholar] [CrossRef] [Green Version] - Margolin, A.A.; Nemenman, I.; Basso, K.; Wiggins, C.; Stolovitzky, G.; Favera, R.D.; Califano, A. ARACNE: An algorithm for the reconstruction of gene regulatory networks in a mammalian cellular context. BMC Bioinform.
**2006**, 7, S7. [Google Scholar] [CrossRef] [Green Version] - Faith, J.J.; Hayete, B.; Thaden, J.T.; Mogno, I.; Wierzbowski, J.; Cottarel, G.; Kasif, S.; Collins, J.J.; Gardner, T.S. Large-scale mapping and validation of Escherichia coli transcriptional regulation from a compendium of expression profiles. PLoS Biol.
**2007**, 5, e8. [Google Scholar] [CrossRef] - Serra, A.; Coretto, P.; Fratello, M.; Tagliaferri, R. Robust and sparse correlation matrix estimation for the analysis of high-dimensional genomics data. Bioinformatics
**2018**, 34, 625–634. [Google Scholar] [CrossRef] - Zhang, X.; Zhao, J.; Hao, J.-K.; Zhao, X.-M.; Chen, L. Conditional mutual inclusive information enables accurate quantification of associations in gene regulatory networks. Nucleic Acids Res.
**2015**, 43, e31. [Google Scholar] [CrossRef] [Green Version] - Zhao, J.; Zhou, Y.; Zhang, X.; Chen, L. Part mutual information for quantifying direct associations in networks. Proc. Natl. Acad. Sci. USA
**2016**, 113, 5130–5135. [Google Scholar] [CrossRef] [Green Version] - Chan, T.E.; Stumpf, M.P.H.; Babtie, A.C. Gene Regulatory Network Inference from Single-Cell Data Using Multivariate Information Measures. Cell Syst.
**2017**, 5, 251–267. [Google Scholar] [CrossRef] [Green Version] - Matsumoto, H.; Kiryu, H.; Furusawa, C.; Ko, M.S.H.; Ko, S.B.H.; Gouda, N.; Hayashi, T.; Nikaido, I. SCODE: An efficient regulatory network inference algorithm from single-cell RNA-Seq during differentiation. Bioinformatics
**2017**, 33, 2314–2321. [Google Scholar] [CrossRef] - Frankowski, P.C.A.; Vert, J.P. Gene regulation inference from single-cell RNA-seq data with linear differential equations and velocity inference. Bioinformatics
**2020**, 36, 4774–4780. [Google Scholar] [CrossRef] - Ma, B.; Fang, M.; Jiao, X. Inference of gene regulatory networks based on nonlinear ordinary differential equations. Bioinformatics
**2020**, 36, 4885–4893. [Google Scholar] [CrossRef] - Herrera-Delgado, E.; Perez-Carrasco, R.; Briscoe, J.; Sollich, P. Memory functions reveal structural properties of gene regulatory networks. PLoS Comput. Biol.
**2018**, 14, e1006003. [Google Scholar] [CrossRef] - Tian, T.; Burrage, K. Stochastic models for regulatory networks of the genetic toggle switch. Proc. Natl. Acad. Sci. USA
**2006**, 103, 8372–8377. [Google Scholar] [CrossRef] [Green Version] - Barman, S.; Kwon, Y.-K. A Boolean network inference from time-series gene expression data using a genetic algorithm. Bioinformatics
**2018**, 34, i927–i933. [Google Scholar] [CrossRef] [Green Version] - Zhang, R.; Ren, Z.; Chen, W. SILGGM: An extensive R package for efficient statistical inference in large-scale gene networks. PLoS Comput. Biol.
**2018**, 14, e1006369. [Google Scholar] [CrossRef] [PubMed] - Friedman, N. Inferring Cellular Networks Using Probabilistic Graphical Models. Science
**2004**, 303, 799–805. [Google Scholar] [CrossRef] [PubMed] - Dimitrakopoulou, K.; Tsimpouris, C.; Papadopoulos, G.; Pommerenke, C.; Wilk, E.; Sgarbas, K.N.; Schughart, K.; Bezerianos, A. Dynamic gene network reconstruction from gene expression data in mice after influenza A (H1N1) infection. J. Clin. Bioinforma.
**2011**, 1, 27. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Xing, L.; Guo, M.; Liu, X.; Wang, C.; Zhang, L. Gene Regulatory Networks Reconstruction Using the Flooding-Pruning Hill-Climbing Algorithm. Genes
**2018**, 9, 342. [Google Scholar] [CrossRef] [Green Version] - Staunton, P.M.; Miranda-Casoluengo, A.A.; Loftus, B.J.; Gormley, I.C. BINDER: Computationally inferring a gene regulatory network for Mycobacterium abscessus. BMC Bioinform.
**2019**, 20, 466. [Google Scholar] [CrossRef] - Magnusson, R.; Gustafsson, M. LiPLike: Towards gene regulatory network predictions of high certainty. Bioinformatics
**2020**, 36, 2522–2529. [Google Scholar] [CrossRef] [Green Version] - Omranian, N.; Eloundou-Mbebi, J.M.O.; Mueller-Roeber, B.; Nikoloski, Z. Gene regulatory network inference using fused LASSO on multiple data sets. Sci. Rep.
**2016**, 6, 20533. [Google Scholar] [CrossRef] [Green Version] - Ghosh Roy, G.; Geard, N.; Verspoor, K.; He, S. PoLoBag: Polynomial Lasso Bagging for signed gene regulatory network inference from expression data. Bioinformatics
**2021**, 36, 5187–5193. [Google Scholar] [CrossRef] - Haury, A.C.; Mordelet, F.; Vera-Licona, P.; Vert, J.P. TIGRESS: Trustful Inference of Gene REgulation using Stability Selection. BMC Syst. Biol.
**2012**, 6, 145. [Google Scholar] [CrossRef] [Green Version] - Mordelet, F.; Vert, J.P. SIRENE: Supervised inference of regulatory networks. Bioinformatics
**2008**, 24. [Google Scholar] [CrossRef] - Huynh-Thu, V.A.; Irrthum, A.; Wehenkel, L.; Geurts, P. Inferring regulatory networks from expression data using tree-based methods. PLoS ONE
**2010**, 5, e12776. [Google Scholar] [CrossRef] - Huynh-Thu, V.A.; Sanguinetti, G. Combining tree-based and dynamical systems for the inference of gene regulatory networks. Bioinformatics
**2015**, 31, 1614–1622. [Google Scholar] [CrossRef] [Green Version] - Moerman, T.; Aibar Santos, S.; Bravo González-Blas, C.; Simm, J.; Moreau, Y.; Aerts, J.; Aerts, S. GRNBoost2 and Arboreto: Efficient and scalable inference of gene regulatory networks. Bioinformatics
**2019**, 35, 2159–2161. [Google Scholar] [CrossRef] - Zheng, R.; Li, M.; Chen, X.; Wu, F.X.; Pan, Y.; Wang, J. BiXGBoost: A scalable, flexible boosting-based method for reconstructing gene regulatory networks. Bioinformatics
**2019**, 35, 1893–1900. [Google Scholar] [CrossRef] - Maraziotis, I.A.; Dragomir, A.; Thanos, D. Gene regulatory networks modelling using a dynamic evolutionary hybrid. BMC Bioinformatics
**2010**, 11, 140. [Google Scholar] [CrossRef] [Green Version] - Yang, Y.; Fang, Q.; Shen, H. Bin Predicting gene regulatory interactions based on spatial gene expression data and deep learning. PLoS Comput. Biol.
**2019**, 15, e1007324. [Google Scholar] [CrossRef] - Penfold, C.A.; Millar, J.B.A.; Wild, D.L. Inferring orthologous gene regulatory networks using interspecies data fusion. Bioinformatics
**2015**, 31, i97–i105. [Google Scholar] [CrossRef] [Green Version] - Noor, A.; Serpedin, E.; Nounou, M.; Nounou, H.; Mohamed, N.; Chouchane, L. An overview of the statistical methods used for inferring gene regulatory networks and protein-protein interaction networks. Adv. Bioinform.
**2013**, 2013, 953814. [Google Scholar] [CrossRef] [Green Version] - Lecca, P.; Priami, C. Biological network inference for drug discovery. Drug Discov. Today
**2013**, 18, 256–264. [Google Scholar] [CrossRef] - Emmert-Streib, F.; Dehmer, M.; Haibe-Kains, B. Gene regulatory networks and their applications: Understanding biological and medical problems in terms of networks. Front. Cell Dev. Biol.
**2014**, 2, 38. [Google Scholar] [CrossRef] [Green Version] - Muldoon, J.J.; Yu, J.S.; Fassia, M.K.; Bagheri, N. Network inference performance complexity: A consequence of topological, experimental and algorithmic determinants. Bioinformatics
**2019**, 35, 3421–3432. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Maetschke, S.R.; Madhamshettiwar, P.B.; Davis, M.J.; Ragan, M.A. Supervised, semi-supervised and unsupervised inference of gene regulatory networks. Brief. Bioinform.
**2014**, 15, 195–211. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Studham, M.E.; Tjärnberg, A.; Nordling, T.E.M.; Nelander, S.; Sonnhammer, E.L.L. Functional association networks as priors for gene regulatory network inference. Bioinformatics
**2014**, 30, i130–i138. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Siahpirani, A.F.; Roy, S. A prior-based integrative framework for functional transcriptional regulatory network inference. Nucleic Acids Res.
**2017**, 45, e21. [Google Scholar] [CrossRef] [Green Version] - Wang, Y.; Goh, W.; Wong, L.; Montana, G. Random forests on Hadoop for genome-wide association studies of multivariate neuroimaging phenotypes. BMC Bioinform.
**2013**, 14 (Suppl. 1), S6. [Google Scholar] [CrossRef] [Green Version] - Wuchty, S.; Arjona, D.; Li, A.; Kotliarov, Y.; Walling, J.; Ahn, S.; Zhang, A.; Maric, D.; Anolik, R.; Zenklusen, J.C.; et al. Prediction of associations between microRNAs and gene expression in glioma biology. PLoS ONE
**2011**, 6, e14681. [Google Scholar] [CrossRef] [Green Version] - Dimitrakopoulos, G.N.; Balomenos, P.; Vrahatis, A.G.; Sgarbas, K.; Bezerianos, A. Identifying disease network perturbations through regression on gene expression and pathway topology analysis. In Proceedings of the 2016 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Orlando, FL, USA, 16–20 August 2016; pp. 5969–5972. [Google Scholar] [CrossRef]
- Dimitrakopoulos, G.N.; Vrahatis, A.G.; Plagianakos, V.; Sgarbas, K. Pathway analysis using XGBoost classification in Biomedical Data. In Proceedings of the 10th Hellenic Conference on Artificial Intelligence (SETN ’18), Patras, Greece, 9–12 July 2018; ACM: New York, NY, USA, 2018; p. 46. [Google Scholar] [CrossRef]
- Dimitrakopoulos, G.N.; Dimitrakopoulou, K.; Maraziotis, I.A.; Sgarbas, K.; Bezerianos, A. Supervised method for construction of microRNA-mRNA networks: Application in cardiac tissue aging dataset. In Proceedings of the 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Chicago, IL, USA, 26–30 August 2014; pp. 318–321. [Google Scholar] [CrossRef]
- Chen, T.; Guestrin, C. XGBoost: A scalable tree boosting system. In Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA, 13–17 August 2016; ACM: New York, NY, USA, 2016; pp. 785–794. [Google Scholar] [CrossRef] [Green Version]
- Friedman, J.H. Greedy function approximation: A gradient boosting machine. Ann. Stat.
**2001**, 29, 1189–1232. [Google Scholar] [CrossRef] - Marbach, D.; Prill, R.J.; Schaffter, T.; Mattiussi, C.; Floreano, D.; Stolovitzky, G. Revealing strengths and weaknesses of methods for gene network inference. Proc. Natl. Acad. Sci. USA
**2010**, 107, 6286–6291. [Google Scholar] [CrossRef] [Green Version] - Marbach, D.; Costello, J.C.; Küffner, R.; Vega, N.M.; Prill, R.J.; Camacho, D.M.; Allison, K.R.; Kellis, M.; Collins, J.J.; Aderhold, A.; et al. Wisdom of crowds for robust gene network inference. Nat. Methods
**2012**, 9, 796–804. [Google Scholar] [CrossRef] [Green Version] - Lawlor, N.; George, J.; Bolisetty, M.; Kursawe, R.; Sun, L.; Sivakamasundari, V.; Kycia, I.; Robson, P.; Stitzel, M.L. Single-cell transcriptomes identify human islet cell signatures and reveal cell-type-specific expression changes in type 2 diabetes. Genome Res.
**2017**, 2, 208–222. [Google Scholar] [CrossRef] - Madhamshettiwar, P.B.; Maetschke, S.R.; Davis, M.J.; Reverter, A.; Ragan, M.A. Gene regulatory network inference: Evaluation and application to ovarian cancer allows the prioritization of drug targets. Genome Med.
**2012**, 4, 41. [Google Scholar] [CrossRef] [Green Version] - Espinosa-Soto, C. On the role of sparseness in the evolution of modularity in gene regulatory networks. PLoS Comput. Biol.
**2018**, 14, e1006172. [Google Scholar] [CrossRef] - Ouma, W.Z.; Pogacar, K.; Grotewold, E. Topological and statistical analyses of gene regulatory networks reveal unifying yet quantitatively different emergent properties. PLoS Comput. Biol.
**2018**, 14, e1006098. [Google Scholar] [CrossRef] - Chen, S.; Mar, J.C. Evaluating methods of inferring gene regulatory networks highlights their lack of performance for single cell gene expression data. BMC Bioinform.
**2018**, 19, 232. [Google Scholar] [CrossRef] [Green Version] - Raser, J.M.; O’Shea, E.K. Noise in gene expression: Origins, consequences, and control. Science
**2005**, 309, 2010–2013. [Google Scholar] [CrossRef] [Green Version] - Tarca, A.L.; Draghici, S.; Khatri, P.; Hassan, S.S.; Mittal, P.; Kim, J.S.; Kim, C.J.; Kusanovic, J.P.; Romero, R. A novel signaling pathway impact analysis. Bioinformatics
**2009**, 24, 75–82. [Google Scholar] [CrossRef] [Green Version] - Vrahatis, A.G.; Balomenos, P.; Tsakalidis, A.K.; Bezerianos, A. DEsubs: An R package for flexible identification of differentially expressed subpathways using RNA-seq experiments. Bioinformatics
**2016**, 32, 3844–3846. [Google Scholar] [CrossRef] [Green Version] - Judeh, T.; Johnson, C.; Kumar, A.; Zhu, D. TEAK: Topology Enrichment Analysis frameworK for detecting activated biological subpathways. Nucleic Acids Res.
**2013**, 41, 1425–1437. [Google Scholar] [CrossRef] [Green Version] - Vrahatis, A.G.; Dimitrakopoulos, G.N.; Tsakalidis, A.K.; Bezerianos, A. Identifying miRNA-mediated signaling subpathways by integrating paired miRNA/mRNA expression data with pathway topology. In Proceedings of the 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Milan, Italy, 25–29 August 2015; pp. 3997–4000. [Google Scholar] [CrossRef]

**Figure 1.**The workflow of the XGRN method. (

**a**) The input consists of known interactions and a gene expression dataset. Next, using the directed A–B interaction, we predict a score for the A–D interaction. (

**b**) An XGBoost regression model is trained for each known interaction using the corresponding expression profiles. As input is set, the target’s expression and as output the regulator’s expression. (

**c**) Using the trained model with the expression of gene D as input, a score is provided for interaction A–D based on R

^{2}measurement comparing the predicted gene expression A’ and the actual gene expression A.

**Figure 2.**(

**a**) An XGBoost regression model is trained based on a known TF-target gene interaction using the expression profile of the target gene as input (blue) and the TF’s profile as output (orange). Then, the model was used in testing with input the profile of a second candidate gene (green), which was a target of this TF. The prediction (red) matched the profile of the TF (R

^{2}= 0.81); thus, this gene was considered as a target. (

**b**) Similarly, the same model was tested on another candidate gene (green), which was not the target of this TF. The prediction (red) presented low similarity with TF (orange) (R

^{2}= 0.23); thus, this candidate interaction was considered false.

**Figure 3.**(

**a**) Performance in terms of AUROC of XGBoost when varying the number of estimators (trees) and the learning rate (LR) parameters. The results are the average of the five DREAM 4 datasets. (

**b**) ROC curves of DREAM 4 datasets using XGBoost with LR = 0.05 and 50 trees.

**Figure 4.**Performance of XGRN, SIRENE and unsupervised methods on DREAM 5 datasets. The best unsupervised method was GENIE3 for D51, Pearson’s correlation coefficient for D52, two-way ANOVA for D53 and a correlation-based meta-predictor for D54.

Dataset | Samples | Genes | TFs | Interactions ^{1} | Organism |
---|---|---|---|---|---|

DREAM 4.1 (D41) | 100 | 100 | 100 | 176 | Synthetic |

DREAM 4.2 (D42) | 100 | 100 | 100 | 249 | Synthetic |

DREAM 4.3 (D43) | 100 | 100 | 100 | 195 | Synthetic |

DREAM 4.4 (D44) | 100 | 100 | 100 | 211 | Synthetic |

DREAM 4.5 (D45) | 100 | 100 | 100 | 193 | Synthetic |

DREAM 5.1 (D51) | 487 | 1643 | 178 | 4012 | Synthetic |

DREAM 5.2 (D52) | 53 | 2810 | 38 | 515 | S. aureus |

DREAM 5.3 (D53) | 487 | 4511 | 141 | 2066 | E. coli |

DREAM 5.4 (D54) | 321 | 5950 | 114 | 3940 | S. cerevisiae |

GSE86469 | 638 | 2287 | 280 | 6289 | Human |

^{1}Note: true interactions were provided by DREAM in the respective datasets, while for GSE86469, a list of TF and target genes was used from [56].

Dataset | R^{2} | MSE | MAE |
---|---|---|---|

D41 | 0.7949 | 0.8107 | 0.8098 |

D42 | 0.8406 | 0.8125 | 0.8191 |

D43 | 0.8023 | 0.7890 | 0.7979 |

D44 | 0.7767 | 0.7704 | 0.7692 |

D45 | 0.7747 | 0.7563 | 0.7562 |

Average | 0.7978 | 0.7878 | 0.7905 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2021 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Dimitrakopoulos, G.N.
XGRN: Reconstruction of Biological Networks Based on Boosted Trees Regression. *Computation* **2021**, *9*, 48.
https://doi.org/10.3390/computation9040048

**AMA Style**

Dimitrakopoulos GN.
XGRN: Reconstruction of Biological Networks Based on Boosted Trees Regression. *Computation*. 2021; 9(4):48.
https://doi.org/10.3390/computation9040048

**Chicago/Turabian Style**

Dimitrakopoulos, Georgios N.
2021. "XGRN: Reconstruction of Biological Networks Based on Boosted Trees Regression" *Computation* 9, no. 4: 48.
https://doi.org/10.3390/computation9040048