Universally Sloppy Parameter Sensitivities in Systems Biology Models

By Ryan N. Gutenkunst, J. Waterfall, Fergal P. Casey, Kevin S. Brown, Christopher R. Myers, James P. Sethna. PLoS Comput. Biol., 2007

1,336 citations recorded by the source corpus.

Quantitative computational models play an increasingly important role in modern biology. Such models typically involve many free parameters, and assigning their values is often a substantial obstacle to model development. Directly measuring in vivo biochemical parameters is difficult, and collectively fitting them to other experimental data often yields large parameter uncertainties. Nevertheless, in earlier work we showed in a growth-factor-signaling model that collective fitting could yield well-constrained predictions, even when it left individual parameters very poorly constrained. We also showed that the model had a “sloppy” spectrum of parameter sensitivities, with eigenvalues roughly evenly distributed over many decades. Here we use a collection of models from the literature to test whether such sloppy spectra are common in systems biology. Strikingly, we find that every model we examine has a sloppy spectrum of sensitivities. We also test several consequences of this sloppiness for building predictive models. In particular, sloppiness suggests that collective fits to even large amounts of ideal time-series data will often leave many parameters poorly constrained. Tests over our model collection are consistent with this suggestion. This difficulty with collective fits may seem to argue for direct parameter measurements, but sloppiness also implies that such measurements must be formidably precise and complete to usefully constrain many model predictions. We confirm this implication in our growth-factor-signaling model. Our results suggest that sloppy sensitivity spectra are universal in systems biology models. The prevalence of sloppiness highlights the power of collective fits and suggests that modelers should focus on predictions rather than on parameters.

Subfields: Genetics, Molecular Biology

View Paper: https://www.semanticscholar.org/paper/38365ace2637b665eecb6eaa1cf4fb036cfa18d9

How researchers see this paper

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

In Genetics

In Molecular Biology

Citations in the corpus

Approximate Bayesian computation scheme for parameter inference and model selection in dynamical systems (2009, background): “In other words, the smaller PCs correspond to stiﬀ parameter combinations, while the larger PCs may correspond to sloppy parameter combinations (Gutenkunst et al. , 2007).”
Sensitivity analysis approaches applied to systems biology models. (2011, background): “They found that systems biology models exhibit sloppy sensitivity spectra because the behaviour of the models is very insensitive to many parameter changes [72].”
Automated adaptive inference of phenomenological dynamical models (2015, neutral): “Further, speciﬁcation of myriads of microscopic mechanistic parameters in such models demands vast data sets and computational resources, and is hard even for very large data sets due to widely varying sensitivities of predictions to the parameters 3 .”
Absolute Metabolite Concentrations and Implied Enzyme Active Site Occupancy in Escherichia coli (2009, background): “This facilitates otherwise intractable parameter identification challenges in building dynamic Michaelis-Menten models of metabolis”
Parameter Estimation and Model Selection in Computational Biology (2010, background): “Such ‘‘sloppiness’’ has been suggested as a factor that makes parameter estimation difficult [28].”
A fast, robust, and tunable synthetic gene oscillator (2008, background): “This approach was chosen to avoid preferential treatment of a particular set of the 21 varied parameters (Gutenkunst et al. 2007): k`, ku`, ba, br, α, ta, tr, k f a, k f r, kda, k−da, kdr, k−dr, kt, k−t, da, dr, γ, ce, e, λ.”
Recent Developments in Parameter Estimation and Structure Identification of Biochemical and Genomic Systems (2009, background): “some articles [130,132,133], but will require more work.”
Input–output behavior of ErbB signaling pathways as revealed by a mass action model trained against dynamic data (2009, background): “Why then, do we not simply double or triple the amount of data used for model training? Unfortunately, it has been shown that even a complete set of time-resolved measurements of all species in a complex reaction model is insufficient to fully constrain all parameters due to the ‘sloppy’ nature of t”