The results of our study will be made available to the protein structure, bioinformatics and computer science communities by placing on a Web server the database of protein geometry together with the learnt rules in the syntax of both PROLOG clauses and as a ASCII text. Data for a given protein could readily be extracted. Rules could be highlighted visually on a particular protein by launching the RASMOL graphics program. We also plan to explore developing a server to which coordinates will be sent, the PROLOG clauses for the protein structure generated and this compared to the current PROLOG database of protein structure and learnt rules. This could be implemented by an interprocessor communication line between a JAWA program and a PROLOG compiler. Such technology may well be readily available in the third year of the project facilitating this implementation.