FAMoS: A Flexible and dynamic Algorithm for Model Selection to analyse complex systems dynamics

Dynamic contrast enhanced MRI of pulmonary adenocarcinomas for early risk stratification: higher contrast uptake associated with response and better prognosis

BACKGROUND

To explore the prognostic value of serial dynamic contrast-enhanced (DCE) MRI in patients with advanced pulmonary adenocarcinoma undergoing first-line therapy with either tyrosine-kinase inhibitors (TKI) or platinum-based chemotherapy (PBC).

METHODS

Patients underwent baseline (day 0, n = 98), and post-therapeutic DCE MRI (PBC: day + 1, n = 52); TKI: day + 7, n = 46) at 1.5T. Perfusion curves were acquired at 10, 40, and 70 s after contrast application and analysed semiquantitatively. Treatment response was evaluated at 6 weeks by CT (RECIST 1.1); progression-free survival (PFS) and overall survival  were analysed with respect to clinical and perfusion parameters. Relative uptake was defined as signal difference between contrast and non-contrast images, divided by the non-contrast signal. Predictors of survival were selected using Cox regression analysis. Median follow-up was 825 days.

RESULTS

In pre-therapeutic and early post-therapeutic MRI, treatment responders (n = 27) showed significantly higher relative contrast uptake within the tumor at 70 s after application as compared to non-responders (n = 71, p ≤ 0.02), response defined as PR by RECIST 1.1 at 6 weeks. There was no significant change of perfusion at early MRI after treatment. In multivariate regression analysis of selected parameters, the strongest association with PFS were relative uptake at 40 s in the early post-treatment MRI and pre-treatment clinical data (presence of liver metastases, ECOG performance status).

CONCLUSION

Higher contrast uptake within the tumor at pre-treatment and early post-treatment MRI was associated with treatment response and better prognosis. DCE MRI of pulmonary adenocarcinoma may provide important prognostic information.

The partition representation of enzymatic reaction networks and its application for searching bi-stable reaction systems

The signal transduction system, which is known as a regulatory mechanism for biochemical reaction systems in the cell, has been the subject of intensive research in recent years, and its design methods have become necessary from the viewpoint of synthetic biology. We proposed the partition representation of enzymatic reaction networks consisting of post-translational modification reactions such as phosphorylation, which is an important basic component of signal transduction systems, and attempted to find enzymatic reaction networks with bistability to demonstrate the effectiveness of the proposed representation method. The partition modifiers can be naturally introduced into the partition representation of enzymatic reaction networks when applied to search. By randomly applying the partition modifiers as appropriate, we searched for bistable and resettable enzymatic reaction networks consisting of four post-translational modification reactions. The proposed search algorithm worked well and we were able to find various bistable enzymatic reaction networks, including a typical bistable enzymatic reaction network with positive auto-feedbacks and mutually negative regulations. Since the search algorithm is divided into an evaluation function specific to the characteristics of the enzymatic reaction network to be searched and an independent algorithm part, it may be applied to search for dynamic properties such as biochemical adaptation, the ability to reset the biochemical state after responding to a stimulus, by replacing the evaluation function with one for other characteristics.

COVID-19 pneumonia and its lookalikes: How radiologists perform in differentiating atypical pneumonias

Purpose

To examine the performance of radiologists in differentiating COVID-19 from non-COVID-19 atypical pneumonia and to perform an analysis of CT patterns in a study cohort including viral, fungal and atypical bacterial pathogens.

Methods

Patients with positive RT-PCR tests for COVID-19 pneumonia (n = 90) and non-COVID-19 atypical pneumonia (n = 294) were retrospectively included. Five radiologists, blinded to the pathogen test results, assessed the CT scans and classified them as COVID-19 or non-COVID-19 pneumonia. For both groups specific CT features were recorded and a multivariate logistic regression model was used to calculate their ability to predict COVID-19 pneumonia.

Results

The radiologists differentiated between COVID-19 and non-COVID-19 pneumonia with an overall accuracy, sensitivity, and specificity of 88% ± 4 (SD), 79% ± 6 (SD), and 90% ± 6 (SD), respectively. The percentage of correct ratings was lower in the early and late stage of COVID-19 pneumonia compared to the progressive and peak stage (68 and 71% vs 85 and 89%). The variables associated with the most increased risk of COVID-19 pneumonia were band like subpleural opacities (OR 5.55, p < 0.001), vascular enlargement (OR 2.63, p = 0.071), and subpleural curvilinear lines (OR 2.52, p = 0.021). Bronchial wall thickening and centrilobular nodules were associated with decreased risk of COVID-19 pneumonia with OR of 0.30 (p = 0.013) and 0.10 (p < 0.001), respectively.

Conclusions

Radiologists can differentiate between COVID-19 and non-COVID-19 atypical pneumonias at chest CT with high overall accuracy, although a lower performance was observed in the early and late stage of COVID 19 pneumonia. Specific CT features might help to make the correct diagnosis.

Parallel Tempering with Lasso for model reduction in systems biology

Systems Biology models reveal relationships between signaling inputs and observable molecular or cellular behaviors. The complexity of these models, however, often obscures key elements that regulate emergent properties. We use a Bayesian model reduction approach that combines Parallel Tempering with Lasso regularization to identify minimal subsets of reactions in a signaling network that are sufficient to reproduce experimentally observed data. The Bayesian approach finds distinct reduced models that fit data equivalently. A variant of this approach that uses Lasso to perform selection at the level of reaction modules is applied to the NF-κB signaling network to test the necessity of feedback loops for responses to pulsatile and continuous pathway stimulation. Taken together, our results demonstrate that Bayesian parameter estimation combined with regularization can isolate and reveal core motifs sufficient to explain data from complex signaling systems.

Cells respond to diverse environmental cues using complex networks of interacting proteins and other biomolecules. Mathematical and computational models have become invaluable tools to understand these networks and make informed predictions to rationally perturb cell behavior. However, the complexity of detailed models that try to capture all known biochemical elements of signaling networks often makes it difficult to determine the key regulatory elements that are responsible for specific cell behaviors. Here, we present a Bayesian computational approach, PTLasso, to automatically extract minimal subsets of detailed models that are sufficient to explain experimental data. The method simultaneously calibrates and reduces models, and the Bayesian approach samples globally, allowing us to find alternate mechanistic explanations for the data if present. We demonstrate the method on both synthetic and real biological data and show that PTLasso is an effective method to isolate distinct parts of a larger signaling model that are sufficient for specific data.

TopoFilter: a MATLAB package for mechanistic model identification in systems biology

Background

To develop mechanistic dynamic models in systems biology, one often needs to identify all (or minimal) representations of the biological processes that are consistent with experimental data, out of a potentially large set of hypothetical mechanisms. However, a simple enumeration of all alternatives becomes quickly intractable when the number of model parameters grows. Selecting appropriate dynamic models out of a large ensemble of models, taking the uncertainty in our biological knowledge and in the experimental data into account, is therefore a key current problem in systems biology.

Results

The TopoFilter package addresses this problem in a heuristic and automated fashion by implementing the previously described topological filtering method for Bayesian model selection. It includes a core heuristic for searching the space of submodels of a parametrized model, coupled with a sampling-based exploration of the parameter space. Recent developments of the method allow to balance exhaustiveness and speed of the model space search, to efficiently re-sample parameters, to parallelize the search, and to use custom scoring functions. We use a theoretical example to motivate these features and then demonstrate TopoFilter’s applicability for a yeast signaling network with more than 250’000 possible model structures.

Conclusions

TopoFilter is a flexible software framework that makes Bayesian model selection and reduction efficient and scalable to network models of a complexity that represents contemporary problems in, for example, cell signaling. TopoFilter is open-source, available under the GPL-3.0 license at https://gitlab.com/csb.ethz/TopoFilter. It includes installation instructions, a quickstart guide, a description of all package options, and multiple examples.