Cyclic Causal Discovery from Continuous Equilibrium Data

By Joris M. Mooij, Tom M. Heskes. 2013

95 citations recorded by the source corpus.

We propose a method for learning cyclic causal models from a combination of observational and interventional equilibrium data. Novel aspects of the proposed method are its ability to work with continuous data (without assuming linearity) and to deal with feedback loops. Within the context of biochemical reactions, we also propose a novel way of modeling interventions that modify the activity of compounds instead of their abundance. For computational reasons, we approximate the nonlinear causal mechanisms by (coupled) local linearizations, one for each experimental condition. We apply the method to reconstruct a cellular signaling network from the flow cytometry data measured by Sachs et al. (2005). We show that our method finds evidence in the data for feedback loops and that it gives a more accurate quantitative description of the data at comparable model complexity.

Subfields: Artificial Intelligence, Cognitive Neuroscience, Molecular Biology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Cognitive Neuroscience

In Molecular Biology

Citations in the corpus

Distinguishing Cause from Effect Using Observational Data: Methods and Benchmarks (2014, background): “S. Shimizu, P. O. Hoyer, A. Hyvärinen, and A. J. Kerminen.”
CAM: Causal Additive Models, high-dimensional order search and penalized regression (2013, background): “Finally, it would be very interesting to extend (sparse) permutation search to (possibly nonidentifiable) models with hidden variables [7, 11, 23, 34] or with graph structures allowing for cycles [19, 20, 30, 33].”
UvA-DARE ( Digital Academic Repository ) Causal Consistency of Structural Equation Models (2017, unknown): “[10, 12, 14, 15] and [13] to meaningfully interpret cyclic SEMs.”
Joint Causal Inference from Multiple Contexts (2016, methodology): “, 2015; Forré and Mooij, 2018), (b) methods that pool all data and construct a single context-independent causal graph directly from the pooled data (Cooper, 1997; Cooper and Yoo, 1999; Tian and Pearl, 2001; Sachs et al., 2005; Eaton and Murphy, 2007; Chen et al., 2007; Hauser and Bühlmann, 2012; Mo”
Causal Transfer Learning (2017, background): “…observed in all tasks, except that the target Y = X 2 is unobserved in the test task ( I 1 = 1 ). and Pearl, 2001], soft interventions [Markowetz et al., 2005], fat-hand interventions [Eaton and Murphy, 2007], activity interventions [Mooij and Heskes, 2013], and stochastic versions of all these.”
Structure Learning in Graphical Modeling (2016, methodology): “, 2006; Nandy et al., 2016a). The theoretical properties of hybrid algorithms are less well studied than those of purely score- or constraint-based algorithms. Nandy et al. (2016a) try to fill this gap by studying a simple hybrid algorithm: GES restricted to the search space determined by the condit”
Constraint-based causal discovery from multiple interventions over overlapping variable sets (2014, background): “Mooij and Heskes (2013) propose modelling activity inhibitions by removing outgoing edges of the target variable.”
Estimating the effect of joint interventions from observational data in sparse high-dimensional settings (2014, background): “Activity interventions as considered in [23] can also be represented by mechanism changes.”
Jointly interventional and observational data: estimation of interventional Markov equivalence classes of directed acyclic graphs (2013, background): “However, it has been reported that some experimental conditions affect products of inhibited proteins without changing their own phosphorylation state, and hence rather affect children of intervened variables in our framework rather than the intervention target itself (Ellis and Wong, 2008; Mooij an”
Ancestral Causal Inference (2016, unknown): “This approach conveniently applies to various types of interventions: perfect interventions [16], soft interventions [14], mechanism changes [24], and activity interventions [15].”
Foundations of structural causal models with cycles and latent variables (2016, background): “Other types of interventions, like mechanism changes [72], fat-hand interventions [12], activity interventions [42] and stochastic versions of all these are at least as relevant, but we do not consider them here.”
Graphical criteria for efficient total effect estimation via adjustment in causal linear models (2019, methodology): “Here,  C denotes the consensus graph,  S the acyclic graph estimated by Sachs et al. (2005) and  M the acyclic graph estimated by Mooij and Heskes (2013).”
UvA-DARE (Digital Academic Repository) Methods for causal inference from gene perturbation experiments and validation (2016, unknown): “Edge [8]a [8]b [6] [2] ICP hiddenICP RAF→MEK X X X MEK→RAF X X X MEK→ERK X X X PLCg→PIP2 X X X X X PLCg→PIP3 X X PLCg→PKC X X PIP2→PLCg X X PIP2→PIP3 X PIP2→PKC X PIP3→PLCg X PIP3→PIP2 X X X X X PIP3→AKT X AKT→ERK X X X ERK→AKT X X X X ERK→PKA X PKA→RAF X X PKA→MEK X X X X PKA→ERK X X X PKA→AKT X X ”