Representation of lipid ambiguity

Hi

I am pretty new to the field of cheminformatics, I wondered if anyone has knowledge of how ambiguous structures of lipids can be/are represented and treated. To illustrate what I mean:

take the example of glycerophospholipids, which have a "head group" (such as

1,2-diacyl-sn-glycero-3-phosphocholine, see CHEBI:57643) and two tails (R1 + R2). Often biologists specify the general composition of the tails, but not the precise structure. For instance, they might know that the tails (R1 + R2 in the structure) together have 32 carbons, and no double bonds, and would refer to the entity like this: PC(32:0). However they might not be able to specify if the compound is PC(18:0/14:0) (two tails of 18 and 14 carbons) or PC(16:0/16:0) (two tails of 16 carbons) etc. The precise length of each tail may not be known.

I wondered if anyone out there knows of functionalities or methods which have been developed to treat and store such entities. For instance, a method that allows you to load the base structure of the head group (with Rs), and subsequently specify that together the R1 and R2 groups have a certain number of carbon atoms and double bonds, without specifying exactly what each of the R groups actually are. From that information, a formula and mass could be generated (but of course, no structure).

Thanks and Best Regards, Alan

Hi,

If I understand well in the R1, R2  definition you would like to define the number of carbon atoms, number of double bonds, but without the connectivity information and calculate properties of the molecule. Is this the case?

If so, we don't have a solution for this.

Zsolt, could you please comment what types of sgroups are used in the calculations?

Can this problem be solved somehow with sgroups which are taken into acount during the calculations?

Currently sgroups are not really used in calculations. They are ungroupped or expanded before the calculation is executed.

Zsolt

Hi

thanks both for your replies and the information you provide. Volfi, you have precisely understood the requirements - I would like to be able to specify a class of molecule and the composition of the variable "tail" groups, and calculate some information like formula, mass, charge.

Having looked a little more at existing chemical ontologies and classifications, it seems generally they do not provide identifiers for such ambiguous entities (where say "R1+R2" is known), only references to general classes (head + R groups) or defined structures (where both R1 and R2 are known).

Best Regards, Alan