Glycoinformatics

Glycoinformatics is a relatively new field of bioinformatics that pertains to the study of carbohydrates. It broadly includes (but is not restricted to) database, software, and algorithm development for the study of carbohydrate structures, glycoconjugates, enzymatic carbohydrate synthesis and degradation, as well as carbohydrate interactions. Conventional usage of the term does not currently include the treatment of carbohydrates from the more well-known nutritive aspect.

Complexity

Carbohydrates or "sugars" (this term should not be confused with simple sugars - monosaccharides and disaccharides) as they are generally called,^[1] form the third class of biopolymers, other two being proteins and nucleic acids. Unlike proteins and nucleic acids which are linear, carbohydrates are often branched and extremely complex.^[2] For instance, just four sugars can be strung together to form more than 5 million different types of carbohydrates^[3] or nine different sugars may be assembled into 15 million possible four-sugar-chains.^[4] Despite their repetitive nature, carbohydrates are often considered as the "information poor" molecules. Consequently, bioinformatics on glycome is also very poor.^[5]

Sequence representation

The sequence of branching information in a arabinoxylan molecule.

Owing to the lack of a genetic blue print, carbohydrates do not have a "fixed" sequence. Instead, the sequence is largely determined by the kinetic differences in the enzymes and variations in the biosynthetic micro-environment of the cells.

One of the main constrains in the glycoinformatics is the difficulty of representing sugars in the sequence form especially due to their branching nature.^[6]

The sequence of branching information^[7] in a carbohydrate molecule is represented in the figure.

References