Thanks to the development of high-throughput screening technologies, there’s a massive increase in data on chemical compounds and their biological activities. Meanwhile, there’s an increasing demand for computational tools to facilitate drug design, compound synthesis, and test cycle. As a skillful and reliable supplier in the worldwide, Creative Biolabs pays attention to these crucial tools for the network structure of small molecule drugs and provides graph-based similarity assessments to analyze activity data and create new models used in virtual screening methods.
Molecular similarity is a measurement of symmetry and/or resemblance between two objects. It’s a key concept in drug discovery and based on the conception that structurally similar molecules often have similar properties. Assessment of similarity among chemical compounds has been highly effective in the discovery and development of numerous small molecular drugs.
Fig.1 Concepts to approach and quantify the degree of molecular similarity. (Kundert, 2019)
In particular, two-dimensional (2D) similarity approaches have been quite prevalent because of their simplicity, efficiency, and accuracy. Today, the focus is being shifted toward the development of techniques involving the comparison and representation of three-dimensional (3D) conformation of small molecules. Among the 3D similarity approaches, evaluation of shape similarity is now appealing for its application both in virtual screening and molecular target prediction. A wide range of tools has been introduced to describe the molecular shape and determine their similarity features. The most broadly used methods contain graph-based methods, alignment-based methods, atom distance-based methods, surface-based methods, and others.
Graph-based analysis is based on representing the conformer using graphs. One example of graph-based similarity search is RASCAL (Rapid Similarity Calculation) based on similarity search with iterative graph matching procedure. Similarity search detects the level of similarity and regulates the selection of lead compounds, based on user-defined similarity level and the graph match. Another graph-based matching approach is Reduced Graph (RG) based similarity search. Reduced graph provides information about topological pharmacophore, RG considers the molecular subgroups for comparison which takes part in ligand-receptor protein interaction.
Fig.2 Ligand-based virtual screening using graph-based similarity measure. (Hutter, 2011)
Molecular similarity searching frequently needs an approach by which the level of similarity among small molecules can be quantified. The graph-based methods present the compounds using reduced graphs, which gather atomic substructures together according to related features (e.g. hydrogen-bonding, ring systems, and pharmacophoric features). Furthermore, extended reduced graphs are an extension of the reduced graphs and make specific modifications to better describe the pharmacophoric size, characteristics, and shape of the molecules.
Applying similarity for novel compounds is an important issue in drug design. Here, we have listed some general graph-based chemical similarity measures along with recent developments of similarity criteria, including but not limited to:
For the purpose of virtual screening, the compounds require to be encoded into a computer-readable format that allows comparison on the basis of given similarity criteria, comprising the use of 3D structure, fingerprints, graph-based methods, etc. We also provide subsequent data analysis, molecular modifying, and modeling engineering as requests.
Molecular similarity techniques are commonly exploited to select ideal candidates in the drug discovery industry. These methods are involved in various applications such as molecular screening, molecular clustering, and similarity searching. As a famous expert in the drug screening, Creative Biolabs has a series of versatile ligand-based virtual screening services, especially graph-based similarity assessments, for the improvement and the advancement of drug design, drug repurposing and scaffold hopping. These approaches demonstrate excellent virtual screening performance not only retrospectively but also prospectively. If you’re interested in our services, please don’t hesitate to contact us or directly send us an inquiry.
References
For Research Use Only.
