1. Recent developments on statistical approaches for differential network analysis have started to focus on directed networks and in particular directed acyclic graphs (DAGs) (Ghoshal & Honorio 2019; Y. Wang et al. 2018) as well as graphical models for other data types (T. Cai Li Ma & Xia 2018; He et al. 2019; Kim Liu & Kolar 2019; M. Yu Gupta et al. 2019; S. Zhao et al. 2019). A number of software tools have also been developed that provide tests of differential connectivity based on permutation approaches (Gill Datta & Datta 2014) or by considering differences in marginal associations based on correlations instead of conditional dependencies (Fukushima 2013; McKenzie Katsyv Song Wang & Zhang 2016). While these tools may not have strong theoretical support or may test different hypotheses they provide more convenient user interfaces and may be more computationally amenable for analysis of large networks.
2. A Local Poisson Graphical Model for Inferring Networks From Sequencing Data
3. Structural equation modeling in social science research
4. Model selection through sparse maximum likelihood estimation for multivariate Gaussian or binary data;Banerjee O.;Journal of Machine Learning Research,2008