Factorial(4)=24, works every time.
Use lisfact_2_mod.mol and the script moving_random_metabo.sh from the gh-pages branch of the repo.
By a modification of this example, now I have a working factorial computation. (I.e. it works every time.) Recall that I took a lambda term for the factorial from the lambda calculus tutorial by Mayer Goldberg from the little-lisper.org, par. 57, page 14. Then I modified it and got a molecule which computes the factorial in about 20% of the cases. Now, in this working factorial example, I made two supplementary modifications. The first consists in starting from a lambda term which uses the mutiplication in the form L mnf.m(nf) instead of the one used in the tutorial. Secondly, the result of the computation (i.e. the "value" of the factorial) is applied to a SUCC (successor) which is then applied to c0, which result in the generation of the correct result.