Nanoinformatics

NovaMechanics has developed several nanostructure-activity predictive models based on in silico analyses that significantly reduces the time and cost of future experiments, as well as reduce the use of animals. Nanoinformatics tools and techniques established by NovaMechanics evaluate the correlation between structural/physicochemical properties of a broad range of NMs and their specific biological responses (i.e., toxicity, biological activity) in order to obtain a comprehensive risk evaluation for the NMs (development of robust SAR models).

The company has a great expertise in the following  areas related to nanoinformatics:

Design of Experiments (DoE)

To ensure that nanomaterials datasets are delivered well-organized Design of Experiments (DoE) algorithms applied in order to identify missing data on nanomaterials experimental characterization and biological and environmental impact evaluation. NovaMechanics’ approach is to propose new experiments when missing data are considered as critical for further analysis and/or when it is crucial that the dataset is extended to more nanomaterials. For that, NovaMechanics designs the minimum number of experiments required to fill the gaps that will be identified as critical, to increase the quality and the quantity of the available data and produce more accurate, robust and predictive models and tools.

Identification of crucial structural and physico-chemical properties associated with toxicological and ecotoxicological properties of nanomaterials

NovaMechanics’ algorithms and tools (Enalos+ KNIME nodes, Enalos Suite)  identifies and utilizes new molecular descriptors being either computational or experimentally derived. The main criterion for selection of descriptors will be their relevance to key molecular properties and biological responses that are directly associated with toxicological and ecotoxicological effects.

Image analysis descriptors derived from the structure of nanomaterials

NovaMechanics’ algorithms performed quantitative image analysis to identify important structural properties of nanomaterials from microscopy images (structural nanodescriptors). The correct interpretation of results requires skills and relevant experience as it is very likely to obtain misleading results. NovaMechanics’ scientists’ guaranty that the analysis of the results will be done very carefully, in order to reliably estimate all the significant properties related to the structure of nanomaterials.

Computational analysis of big OMICS nanomaterials data

NovaMechanics develops cutting edge computational and statistical tools to identify candidate molecular mechanisms predictive of biological (animal and human) toxicity pathways, by integrating the diverse set of endpoints, gene expression and metabolomics data collected across different species and cell lines. The computational analysis of OMICS is accelerated by GPU algorithms included in NovaMechanics e-infrastructure.

Quantum-mechanics (QM) descriptors for Nanomaterials

Several studies have demonstrated that high-level quantum mechanical methods, such as DFT (Density Functional Theory) and ab initio Hartree-Fock (HF) are able to calculate properties for NMs containing up to 40 atoms (small clusters), while semi-empirical PM6 can compute properties for systems having tenfold larger number of atoms. Despite the apparent differences in the levels of accuracy between semi-empirical and ab initio approaches, it has been demonstrated that for the estimation of significant nanomaterials properties , quantum mechanics provides results, which are similar to those derived by DFT or ab-initio methods. NovaMechanics’ Quantum Mechanics computational workflows showed that semi-empirical methods can perform calculations for nanomaterials systems containing up to 5000 atoms. Our research is continued in the direction to develop new techniques to handle larger clusters of nanomaterials that can be directly compared with the experimental systems.

 Molecular Dynamics Simulations of Nanoparticles

Molecular dynamics (MD) approaches have been widely used over the years for the simulation of biomolecules, such as proteins, nucleic acids, and lipids. Recently, MD methods are also successfully employed for the dynamic description of nanomaterials. NovaMechanics Molecualer Dynamics (MD) workflows enable the identification of structural properties of the nanoparticles. This is particularly useful for many NM systems whose structural details are not usually accompanied by experimental evidence. Thus, intriguing questions, such as structural changes that govern NM evolution, the degree of NM aggregation in different media, principal interactions between nanomaterials (NMs) and biomolecules (protein corona formation), and the effect of interacting molecules on the structure of NMs could be elucidated, with apparent implications for further research and NM optimization.