1) (written for structural biologists and computational biologists).
This project provides a collection of tools for computing geometric measures of biomolecules such as volumes and surfaces areas and for identifying their internal cavities and pockets, if any, and computing their measures. By focusing on the geometry of biomolecules, AlphaMol provides means to understand the role of shape in biomolecular functions and dynamics.

2) (written for the modeler or developer)
This project is a collection of programs for computing the geometric measures of a union of balls. It includes computation of surface areas and volumes, and their derivatives with respect to the positions of the centers of the balls, as well as the identification of cavities and pockets. It is based on the alpha shape theory. It proceeds by first computing the weighted Delaunay triangulation of the center of the balls, followed by a filtering of its simplices to only keep those corresonding to actual overlaps of the balls, yielding the dual complex. The volume and surface area of the balls are then derived by applying the inclusion-exclusion formula on the simplices of the dual complex. The program was optimized for speed and memory usage: currently, it can handle one million balls with less than 600 MB of memory, and compute their surface areas and volumes in les than 7 minutes on a 2GHz dual core processor. The programs are written in C and Fortran. Calculations are performed using floating point representation; if a geometric primitive becomes inconclusive however, calculation is switched to arbitrary precision arithmetics, using the GMP library.