Florian Martin

Assessment of network perturbation amplitudes by applying high-throughput data to causal biological networks

BackgroundHigh-throughput measurement technologies produce data sets that have the potential to elucidate the biological impact of disease, drug treatment, and environmental agents on humans. The scientific community faces an ongoing challenge in the analysis of these rich data sources to more accurately characterize biological processes that have been perturbed at the mechanistic level. Here, a new approach is built on previous methodologies in which high-throughput data was interpreted using prior biological knowledge of cause and effect relationships. These relationships are structured into network models that describe specific biological processes, such as inflammatory signaling or cell cycle progression. This enables quantitative assessment of network perturbation in response to a given stimulus.ResultsFour complementary methods were devised to quantify treatment-induced activity changes in processes described by network models. In addition, companion statistics were developed to qualify significance and specificity of the results. This approach is called Network Perturbation Amplitude (NPA) scoring because the amplitudes of treatment-induced perturbations are computed for biological network models. The NPA methods were tested on two transcriptomic data sets: normal human bronchial epithelial (NHBE) cells treated with the pro-inflammatory signaling mediator TNFα, and HCT116 colon cancer cells treated with the CDK cell cycle inhibitor R547. Each data set was scored against network models representing different aspects of inflammatory signaling and cell cycle progression, and these scores were compared with independent measures of pathway activity in NHBE cells to verify the approach. The NPA scoring method successfully quantified the amplitude of TNFα-induced perturbation for each network model when compared against NF-κB nuclear localization and cell number. In addition, the degree and specificity to which CDK-inhibition affected cell cycle and inflammatory signaling were meaningfully determined.ConclusionsThe NPA scoring method leverages high-throughput measurements and a priori literature-derived knowledge in the form of network models to characterize the activity change for a broad collection of biological processes at high-resolution. Applications of this framework include comparative assessment of the biological impact caused by environmental factors, toxic substances, or drug treatments.

A network-based approach to quantifying the impact of biologically active substances

Fe at u re s P E R S P E C T IV E Society increasingly demands close scrutiny of the potential health risks of long-term exposure to biologically active substances, such as therapeutic drugs or environmental toxins. Such risks are typically assessed a posteriori through clinical epidemiology studies. However, disease might take decades to manifest, at a point where changes in therapeutic regime, life style or exposure would not prevent disease onset. Moreover, disease risk as assessed correlatively in epidemiological studies is not intended to elucidate the mechanisms that link perturbations in molecular signaling to disease and, thus, provides fewer options for intervention. Here, we propose that network-based approaches to pharmacology are a valuable way to not only quantify biological network perturbations caused by active substances, but also identify mechanisms and biomarkers modulated in response to exposure and related to disease onset. We also discuss progress towards a generalizable approach for a mechanistic biological impact assessment. Novel computational methods that derive the quantitative biological impact [defined as a biological impact factor (BIF)] from underlying system-wide data using defined causal biological (i.e. molecular) network models as the substrate for data analysis are currently under

Strengths and limitations of microarray-based phenotype prediction: lessons learned from the IMPROVER Diagnostic Signature Challenge

MOTIVATION After more than a decade since microarrays were used to predict phenotype of biological samples, real-life applications for disease screening and identification of patients who would best benefit from treatment are still emerging. The interest of the scientific community in identifying best approaches to develop such prediction models was reaffirmed in a competition style international collaboration called IMPROVER Diagnostic Signature Challenge whose results we describe herein. RESULTS Fifty-four teams used public data to develop prediction models in four disease areas including multiple sclerosis, lung cancer, psoriasis and chronic obstructive pulmonary disease, and made predictions on blinded new data that we generated. Teams were scored using three metrics that captured various aspects of the quality of predictions, and best performers were awarded. This article presents the challenge results and introduces to the community the approaches of the best overall three performers, as well as an R package that implements the approach of the best overall team. The analyses of model performance data submitted in the challenge as well as additional simulations that we have performed revealed that (i) the quality of predictions depends more on the disease endpoint than on the particular approaches used in the challenge; (ii) the most important modeling factor (e.g. data preprocessing, feature selection and classifier type) is problem dependent; and (iii) for optimal results datasets and methods have to be carefully matched. Biomedical factors such as the disease severity and confidence in diagnostic were found to be associated with the misclassification rates across the different teams. AVAILABILITY The lung cancer dataset is available from Gene Expression Omnibus (accession, GSE43580). The maPredictDSC R package implementing the approach of the best overall team is available at www.bioconductor.org or http://bioinformaticsprb.med.wayne.edu/.

Causal biological network database: a comprehensive platform of causal biological network models focused on the pulmonary and vascular systems

With the wealth of publications and data available, powerful and transparent computational approaches are required to represent measured data and scientific knowledge in a computable and searchable format. We developed a set of biological network models, scripted in the Biological Expression Language, that reflect causal signaling pathways across a wide range of biological processes, including cell fate, cell stress, cell proliferation, inflammation, tissue repair and angiogenesis in the pulmonary and cardiovascular context. This comprehensive collection of networks is now freely available to the scientific community in a centralized web-based repository, the Causal Biological Network database, which is composed of over 120 manually curated and well annotated biological network models and can be accessed at http://causalbionet.com. The website accesses a MongoDB, which stores all versions of the networks as JSON objects and allows users to search for genes, proteins, biological processes, small molecules and keywords in the network descriptions to retrieve biological networks of interest. The content of the networks can be visualized and browsed. Nodes and edges can be filtered and all supporting evidence for the edges can be browsed and is linked to the original articles in PubMed. Moreover, networks may be downloaded for further visualization and evaluation. Database URL: http://causalbionet.com

Quantification of biological network perturbations for mechanistic insight and diagnostics using two-layer causal models

BackgroundHigh-throughput measurement technologies such as microarrays provide complex datasets reflecting mechanisms perturbed in an experiment, typically a treatment vs. control design. Analysis of these information rich data can be guided based on a priori knowledge, such as networks or set of related proteins or genes. Among those, cause-and-effect network models are becoming increasingly popular and more than eighty such models, describing processes involved in cell proliferation, cell fate, cell stress, and inflammation have already been published. A meaningful systems toxicology approach to study the response of a cell system, or organism, exposed to bio-active substances requires a quantitative measure of dose-response at network level, to go beyond the differential expression of single genes.ResultsWe developed a method that quantifies network response in an interpretable manner. It fully exploits the (signed graph) structure of cause-and-effect networks models to integrate and mine transcriptomics measurements. The presented approach also enables the extraction of network-based signatures for predicting a phenotype of interest. The obtained signatures are coherent with the underlying network perturbation and can lead to more robust predictions across independent studies. The value of the various components of our mathematically coherent approach is substantiated using several in vivo and in vitro transcriptomics datasets. As a proof-of-principle, our methodology was applied to unravel mechanisms related to the efficacy of a specific anti-inflammatory drug in patients suffering from ulcerative colitis. A plausible mechanistic explanation of the unequal efficacy of the drug is provided. Moreover, by utilizing the underlying mechanisms, an accurate and robust network-based diagnosis was built to predict the response to the treatment.ConclusionThe presented framework efficiently integrates transcriptomics data and “cause and effect” network models to enable a mathematically coherent framework from quantitative impact assessment and data interpretation to patient stratification for diagnosis purposes.

Quantitative assessment of biological impact using transcriptomic data and mechanistic network models

Exposure to biologically active substances such as therapeutic drugs or environmental toxicants can impact biological systems at various levels, affecting individual molecules, signaling pathways, and overall cellular processes. The ability to derive mechanistic insights from the resulting system responses requires the integration of experimental measures with a priori knowledge about the system and the interacting molecules therein. We developed a novel systems biology-based methodology that leverages mechanistic network models and transcriptomic data to quantitatively assess the biological impact of exposures to active substances. Hierarchically organized network models were first constructed to provide a coherent framework for investigating the impact of exposures at the molecular, pathway and process levels. We then validated our methodology using novel and previously published experiments. For both in vitro systems with simple exposure and in vivo systems with complex exposures, our methodology was able to recapitulate known biological responses matching expected or measured phenotypes. In addition, the quantitative results were in agreement with experimental endpoint data for many of the mechanistic effects that were assessed, providing further objective confirmation of the approach. We conclude that our methodology evaluates the biological impact of exposures in an objective, systematic, and quantifiable manner, enabling the computation of a systems-wide and pan-mechanistic biological impact measure for a given active substance or mixture. Our results suggest that various fields of human disease research, from drug development to consumer product testing and environmental impact analysis, could benefit from using this methodology.

A 28-day rat inhalation study with an integrated molecular toxicology endpoint demonstrates reduced exposure effects for a prototypic modified risk tobacco product compared with conventional cigarettes.

Towards a systems toxicology-based risk assessment, we investigated molecular perturbations accompanying histopathological changes in a 28-day rat inhalation study combining transcriptomics with classical histopathology. We demonstrated reduced biological activity of a prototypic modified risk tobacco product (pMRTP) compared with the reference research cigarette 3R4F. Rats were exposed to filtered air or to three concentrations of mainstream smoke (MS) from 3R4F, or to a high concentration of MS from a pMRTP. Histopathology revealed concentration-dependent changes in response to 3R4F that were irritative stress-related in nasal and bronchial epithelium, and inflammation-related in the lung parenchyma. For pMRTP, significant changes were seen in the nasal epithelium only. Transcriptomics data were obtained from nasal and bronchial epithelium and lung parenchyma. Concentration-dependent gene expression changes were observed following 3R4F exposure, with much smaller changes for pMRTP. A computational-modeling approach based on causal models of tissue-specific biological networks identified cell stress, inflammation, proliferation, and senescence as the most perturbed molecular mechanisms. These perturbations correlated with histopathological observations. Only weak perturbations were observed for pMRTP. In conclusion, a correlative evaluation of classical histopathology together with gene expression-based computational network models may facilitate a systems toxicology-based risk assessment, as shown for a pMRTP.

Case study: the role of mechanistic network models in systems toxicology

Twenty first century systems toxicology approaches enable the discovery of biological pathways affected in response to active substances. Here, we briefly summarize current network approaches that facilitate the detailed mechanistic understanding of the impact of a given stimulus on a biological system. We also introduce our network-based method with two use cases and show how causal biological network models combined with computational methods provide quantitative mechanistic insights. Our approach provides a robust comparison of the transcriptional responses in different experimental systems and enables the identification of network-based biomarkers modulated in response to exposure. These advances can also be applied to pharmacology, where the understanding of disease mechanisms and adverse drug effects is imperative for the development of efficient and safe treatment options.

A 7-month cigarette smoke inhalation study in C57BL/6 mice demonstrates reduced lung inflammation and emphysema following smoking cessation or aerosol exposure from a prototypic modified risk tobacco product.

Modified risk tobacco products (MRTP) are designed to reduce smoking-related health risks. A murine model of chronic obstructive pulmonary disease (COPD) was applied to investigate classical toxicology end points plus systems toxicology (transcriptomics and proteomics). C57BL/6 mice were exposed to conventional cigarette smoke (3R4F), fresh air (sham), or a prototypic MRTP (pMRTP) aerosol for up to 7 months, including a cessation group and a switching-to-pMRTP group (2 months of 3R4F exposure followed by fresh air or pMRTP for up to 5 months respectively). 3R4F smoke induced the typical adaptive changes in the airways, as well as inflammation in the lung, associated with emphysematous changes (impaired pulmonary function and alveolar damage). At nicotine-matched exposure concentrations of pMRTP aerosol, no signs of lung inflammation and emphysema were observed. Both the cessation and switching groups showed a similar reversal of inflammatory responses and no progression of initial emphysematous changes. A significant impact on biological processes, including COPD-related inflammation, apoptosis, and proliferation, was identified in 3R4F-exposed, but not in pMRTP-exposed lungs. Smoking cessation or switching reduced these perturbations to near sham-exposed levels. In conclusion, the mouse model indicated retarded disease progression upon cessation or switching to pMRTP which alone had no adverse effects.

Alterations in the sputum proteome and transcriptome in smokers and early-stage COPD subjects.

Chronic obstructive pulmonary disease (COPD) is one of the most prevalent lung diseases. Cigarette smoking is the main risk factor for COPD. In this parallel-group clinical study we investigated to what extent the transitions in a chronic-exposure-to-disease model are reflected in the proteome and cellular transcriptome of induced sputum samples. We selected 60 age- and gender-matched individuals for each of the four study groups: current asymptomatic smokers, smokers with early stage COPD, former smokers, and never smokers. The cell-free sputum supernatant was analyzed by quantitative proteomics and the cellular mRNA fraction by gene expression profiling. The sputum proteome of current smokers clearly reflected the common physiological responses to smoke exposure, including alterations in mucin/trefoil proteins and a prominent xenobiotic/oxidative stress response. The latter response also was observed in the transcriptome, which additionally demonstrated an immune-cell polarization change. The former smoker group showed nearly complete attenuation of these biological effects. Thirteen differentially abundant proteins between the COPD and the asymptomatic smoker group were identified including TIMP1, APOA1, C6orf58, and BPIFB1 (LPLUNC1). In summary, our study demonstrates that sputum profiling can capture the complex and reversible physiological response to cigarette smoke exposure, which appears to be only slightly modulated in early-stage COPD.