A "secret sharing system" permits a secret to be shared among n trustees in such a way that any k of them can recover the secret, but any k-1 have complete uncertainty about it. A linear coding scheme for secret sharing is exhibited which subsumes the polynomial interpolation method proposed by Shamir and can also he viewed as a deterministic version of Blakley's probabilistic method. Bounds on the maximum value of n for a given k and secret size are derived for any system, linear or nonlinear. The proposed scheme achieves the lower bound which, for practical purposes, differs insignificantly from the upper bound. The scheme may be extended to protect several secrets. Methods to protect against deliberate tampering by any of the trustees are also presented.

A new algorithm for identifying three-dimensional configurations of chemical features common to a set of molecules is described. The algorithm scores each configuration based both on the degree to which it is common to the input set and its estimated rarity. The algorithm can be applied to molecules with large (several hundred) conformational models. Results from the application of this algorithm to three data sets are discussed: PAF antagonists, HIV reverse transcriptase inhibitors, and HIV protease inhibitors. Of particular interest is a common configuration identified for a set of HIV reverse transcriptase inhibitors; this configuration is shared by two new, potent inhibitors that were recently described in the literature.

The advent of three-dimensional (3D) molecular database searching motivates this investigation of how best to formulate queries for compounds likely to bind to an enzyme or receptor. 3D queries in the literature generally refer to simple topological features (e.g., nitrogen or phenyl). To better capture the chemist's intent and to find functionally equivalent but structurally diverse compounds, generalized chemical function definitions are proposed for hydrogen bond acceptors/donors, charge centers, and hydrophobes. Use of these function definitions in 5-HT3 antagonist and ACE inhibition queries is shown to identify dramatically more hits capable of forming the hypothesized interactions. Furthermore, false positives that have features inaccessible to the receptor are eliminated. Next, the literahre on intermolecular interaction energy is reviewed to determine what geometric tolerances are chemically reasonable in queries. Finally, it is shown that the commonly used distance constraints poorly distinguish conformers that do and do not superimpose well with receptor features. An altemative, the location constraint, is proposed. Queries for angiotensin I1 antagonism and HLE inhibition are described and used to search 203 000 compounds.

The shape of and the chemical features of a ligand are both critical for biological activity. This paper presents a strategy that uses these descriptors to build a computational model for virtual screening of bioactive compounds. Molecules are represented in a binary shape-feature descriptor space as bit-strings, and their relative activities are used to identify the subset of the bit-string that is most relevant to bioactivity. This subset is used to score virtual libraries. We describe the computational details of the method and present an example validation experiment on thrombin inhibitors.