Many small-molecule databases provide chemical structure search as a free, on-line service. Unfortunately, these services often lack an interoperable interface that conforms to standard protocols, making them inflexible for automated use or combination with other interoperable data sources.
To address these shortcomings, we introduce a publicly accessible service that provides chemical similarity and substructure relations to enhance interoperable searches with additional data about chemical compound structures. This opens up new avenues for searching and querying existing data sources in order to obtain more precise, meaningful data.
Sub-structure/super-structure search
Sub-structure/super-structure search is a relatively recent method for recognizing chemical structures with similarity and/or activity profiles. Unfortunately, the algorithm has some drawbacks, such as its large search space and inflexible matching strategy.
It must be said, the algorithm has been demonstrated successful on a large data set in several experiments. This success can largely be attributed to its innovative backtracking algorithm which applies directly to chemical structure graphs (not association graphs) as well as other clever tricks.
One major drawback of MCS or maximum common subgraph analysis is its difficulty in measuring it. This becomes especially true if the graphs are disconnected, such as when two chemical structures may contain five disconnected C-O fragments. Thus, using MCS might not be the most efficient way to discover a shared substructure among many molecules in an online database; hence it’s essential to select the right algorithm for the job at hand.
Identity search
Identity search is an invaluable feature that permits users to explore account activity data. It also assists in recognizing and resolving any issues with identity-related accounts.
Identity plays a significant role in shaping how one behaves, interacts and makes decisions. Having an authentic sense of self can make life simpler and less stressful for individuals.
People often experience doubts about their sense of identity from time to time, but this is an entirely normal part of life. When faced with major changes or trauma, it can be helpful to speak to someone about how you feel about yourself.
Sometimes our identities can undergo dramatic shifts, such as during adolescence or after becoming parents or moving to a new area. These experiences are known collectively as identity crises.
2-D similarity search
Two-dimensional (2D) similarity search is a widely used approach in drug discovery. This type of search relies on the idea that structurally similar molecules often exhibit similar properties.
Sub-structure search uses the same approach as sub-structure searching, except it looks for replacement fragments instead of searching the target. Furthermore, users can specify required atom or bond features to further narrow their search results.
This type of search can be utilized in several different ways:
One can use this technique to identify compounds similar to Ligase inhibitors that possess a low energy conformer that matches the shape and position of its pharmacophore features.
Another application of this type of search is drug repurposing. Here, one can use a similarity map to untangle large clusters of molecules by modulating an automatically determined similarity threshold based on each molecule’s neighborhood situation.
3-D similarity search
Three-dimensional (3D) similarity search is an extremely powerful and efficient method for virtual screening, capable of filtering millions of molecules quickly.
Shape similarity methods were originally devised from computational geometry concepts in order to represent and compare the shapes of 3D objects. Fast atom distance-based methods are the most popular, but more complex mathematically-based solutions such as SH or VAMS can also be effective at comparing small molecular shapes.
To facilitate faster and more accurate comparisons, shape-based similarity searches were combined with physicochemical properties such as surface charge distribution, polarity and hydrophobicity of molecules. This information can be beneficial in predicting binding features, drug repurposing strategies and scaffold hopping activities.
Discussion about this post