Python Programming for Biology: Bioinformatics and Beyond

Figure 8.1.  A simple hierarchical data model for organising macromolecular

structure data. The example data model for containing the structures of biological

polymer molecules is composed of four object classes. The top of the hierarchy is a

Structure, which represents the 3D conformation of a group of molecules. Each Structure

contains Chain objects, each of which represents a different molecule of a specific type

(DNA, RNA or protein). A Chain contains Residue objects that represent smaller chemical

components that are linked together into a chain, i.e. amino acids or nucleotides, thus

representing the protein or nucleic acid sequence. Each Residue object contains Atom

objects, which are of a specific chemical element type and carry the actual spatial

coordinates.

A typical life cycle of a Python object involves first creating it, then possibly modifying

it, and finally deleting it (or leaving it to cease to exist when a program stops). Our model

will  explicitly  handle  some  deletion  cases,  like  removing  a  Chain  from  a  Structure.

However, it is very tempting to not worry at all about deleting objects, and often you can

get away with this. Sometimes there is simply no need to add deletion functions, because

you  are  only  ever  going  to  explicitly  create  and  modify  objects,  never  remove  them.  To

this end it is handy that Python has a ‘garbage collection’ mechanism, which means that

once you no longer have a handle to an object, it will (eventually) automatically be cleared

away.