Many-body systems are very hard to simulate due to the explosion of parameters with the system size. Quantum computers can help in this task, although one may need scalable systems, something that is out of reach in the short run. An attractive alternative is provided by analog quantum simulators which, even though they are not universal, they can still be tuned to study interesting problems. Atoms in optical lattices seem to be ideally suited for that task. Most of the proposals of such simulators have focused so far on condensed matter or high energy physics problems. In this talk I will show how one can extend the range of problems to other scenarios, especially to quantum chemistry.