Aurélien Naldi

About my research work

My research work was centered on the study of complex biological systems through formal analysis of qualitative dynamical models. I obtained my PhD at Aix-Marseille University in 2009. My last research position was a starting research position (Inria SRP, a type of associate researcher) in the Lifeware team at Inria Saclay - Île-de-France.

Dynamical models, based on state of the art knowledge of the mechanisms underlying biological processes, can be used to decipher the behaviour of these systems. During the modelling process, the comparison between predictions obtained through the dynamical analysis of such models and the observed phenotypes provide valuable information about gaps or inconsistencies in our knowledge. Models capable of reproducing the observed phenotypes can ultimately be used to guide the design of future experiments.

Qualitative models capture many relevant dynamical behaviours despite their lack of quantitative values. The use of non-deterministic simulation strategies yields a large number of alternative dynamical trajectories, which reflects well the complexity of the systems under study. The removal of unknown quantitative parameters enables the definition of models encompassing dozens of components. Qualitative formalisms further provide a framework to develop efficient analytical methods to identify some dynamical properties without performing costly simulations.

My main contributions to this interdisciplinary research area, combine methodological developments, software implementation, and applications to biological systems.

See also: publications - seminars.

Methodological work

I proposed a method for the identification of stable states (fixed points) by transforming the logical functions associated to all model components into stability conditions. The intersection of these conditions then provides the list of all possible stable states. Note that while this method only provides a list of all possible stable states, it does not imply anything about their reachability properties.

My initial implementation used handcrafted decision diagrams, newer implementations use generic constraint solvers such as the clingo ASP solver. This approach has been generalized by others for the identification of stable motifs which provide a good approximation of complex attractors.

Decision Diagrams for the Representation and Analysis of Logical Models of Genetic Networks. A. Naldi, D. Thieffry, C. Chaouiya. Computational Methods in Systems Biology :233--247 (2007).

I also proposed a model reduction method, allowing to remove some manually-selected components while preserving important dynamical properties. We could show in particular that this reduction preserves stable states as well as minimal stable motifs and gives an under-approximation of reachability properties.

The trajectories obtained with the reduced model correspond to trajectories of the complete model in which the reduced components are always updated before the remaining ones. The reduced model can thus also be interpreted as a kinetic refinement of the full model with a simplification similar to the quasi-steady-state approximation for continuous systems.

Dynamically consistent reduction of logical regulatory graphs. A. Naldi, E. Remy, D. Thieffry, C. Chaouiya. Theoretical Computer Science 412:2207--2218 (2011).

I am currently interrested in the definition of multi-scale models of cell populations, starting with the stochastic approximation of their behaviours.

I currently work on the caracterisation of reachability properties which can be derived from stable motifs. I then aim to use this work to derive efficient abstractions of cell populations based on the analysis of isolated cells.

Software tools and standards

Since 2006, I am the main developer of the GINsim software, which has been used for the definition and analysis of a many logical models. This tool provides a graphical interface for model design and a number of analytical tools.

Logical modelling and analysis of cellular regulatory networks with GINsim 3.0. A. Naldi, C. Hernandez, W. Abou-Jaoudé, P. Monteiro, C. Chaouiya, D. Thieffry. Frontiers in Physiology 9:646 (2018).

The number of software tools for qualitative models grew over the years, each of them using its own file format. To overcome the multiplication of formats and enable interoperability between complementary tools, I participated in the definition of the SBML qual extension, an exchange format for qualitative models.

This effort led to the creation of the CoLoMoTo consortium.

SBML qualitative models: a model representation format and infrastructure to foster interactions between qualitative modelling formalisms and tools. C. Chaouiya, D. Bérenguier, S. Keating, A. Naldi, M. van Iersel, N. Rodriguez, A. Dräger, F. Büchel, T. Cokelaer, B. Kowal, B. Wicks, E. Gonçalves, J. Dorier, M. Page, P. Monteiro, A. Kamp, I. Xenarios, H. de Jong, M. Hucka, S. Klamt, D. Thieffry, N. Novère, J. Saez-Rodriguez, T. Helikar. BMC Systems Biology 7:135 (2013).
Cooperative development of logical modelling standards and tools with CoLoMoTo. A. Naldi, P. Monteiro, C. Müssel, t. Consortium for Logical Models and Tools, H. Kestler, D. Thieffry, I. Xenarios, J. Saez-Rodriguez, T. Helikar, C. Chaouiya. Bioinformatics 31:1154-1159 (2015).


The bioLQM toolbox provides command line and programmatic interfaces for the core data structures and algorithms available in GINsim, as well as import/export filters to improve interoperability beyond the tools supporting the SBML qual format directly.

These tools along with many others are integrated in the CoLoMoTo notebook. This modelling platform focuses on reproducible results through the use of Docker images (providing frozen snapshots of a complex software environment) and Jupyter notebooks to build and share complex analysis workflows.

The CoLoMoTo Interactive Notebook: Accessible and Reproducible Computational Analyses for Qualitative Biological Networks. A. Naldi, C. Hernandez, N. Levy, G. Stoll, P. Monteiro, C. Chaouiya, T. Helikar, A. Zinovyev, L. Calzone, S. Cohen-Boulakia, D. Thieffry, L. Paulevé. Frontiers in Physiology 9:680 (2018).

Logical models and data analysis for biological systems

Qualitative models have been applied to a wide range of biological processes. In particular, GINsim has been used to study dozens of models, either through collaborations or by independent researchers. I have been personally involved mainly in the study of cell cycle, and differentiation of T helper cells.

Dynamical analysis of a generic Boolean model for the control of the mammalian cell cycle. A. Faure, A. Naldi, C. Chaouiya, D. Thieffry. Bioinformatics 22:e124--e131 (2006).
Diversity and Plasticity of Th Cell Types Predicted from Regulatory Network Modelling. A. Naldi, J. Carneiro, C. Chaouiya, D. Thieffry. PLoS Computational Biology 6:e1000912 (2010).


During my postdoctoral work in Lausanne (Switzerland), I was involved in the study of genomics data (microarray and ChIP-seq) in the context of metabolic processes.

Genome-Wide Analysis of SREBP1 Activity around the Clock Reveals Its Combined Dependency on Nutrient and Circadian Signals. F. Gilardi, E. Migliavacca, A. Naldi, M. Baruchet, D. Canella, G. Martelot, N. Guex, B. Desvergne. PLoS Genetics 10:e1004155 (2014).
System analysis of cross-talk between nuclear receptors reveals an opposite regulation of the cell cycle by LXR and FXR in human HepaRG liver cells. L. Wigger, C. Casals-Casas, M. Baruchet, K. Trang, S. Pradervand, A. Naldi, B. Desvergne. PLOS ONE 14:e0220894 (2019).


I also worked on the reconstruction of potential causal paths based on known pathways and proteomics (differential phosphorylation using the SILAC approach) data of cancer cell lines with different metastatic status. This work has since been further extended by a PhD student with my former collaborators.

Reconstruction and Signal Propagation Analysis of the Syk Signaling Network in Breast Cancer Cells. A. Naldi, R. Larive, U. Czerwinska, S. Urbach, P. Montcourrier, C. Roy, J. Solassol, G. Freiss, P. Coopman, O. Radulescu. PLoS Computational Biology 13:e1005432 (2017).
Network Reconstruction and Significant Pathway Extraction Using Phosphoproteomic Data from Cancer Cells. M. Buffard, A. Naldi, O. Radulescu, P. Coopman, R. Larive, G. Freiss. Proteomics 19:e1800450 (2019).