It is only used by those with no background in bioinformatics, but expertise in Mathematica, who decide to tackle a bioinformatics problem.
Proprietary software is rarely used in bioinformatics, and rightly so.
The statistical analysis language of choice in bioinformatics is R, which is completely open source, and has an extensive bioinformatics library of programs.
Short answer would probably be Yes, but not be a very large number of people. Every once in awhile I come across some bit of code someone describes in a paper and it is Mathematica or MatLab and I promptly stop reading because neither are widely used. Unless you're really invested in Mathematica it isn't worth learning solely for bioinformatics (you may find other uses for it perhaps) and definitely not for publishing methods...
There are quite a few branches of biology that make heavy use of mathematics and hence where Mathematica is popular. For example ecology, evolution, population genetics, epidemiology. The book A Biologist's Guide to Mathematical Modeling, which is very good and written by expert in the field, has practicals written in Mathematica.
In any case, I don't know Mathematica at all but I feel quite confident in saying that once you get the hang of mathematical modelling then it should be fairly easy to switch to other platforms to implement it. I mean, the difficult bit is to understand the mathematics, the coding part is usually easier!