Bioinformatics Contest 2017 ★ 4.250

Good and creative competition, an excellent exercise for both the experienced and fresh coders. I would only like to point out that some of the exercises are universal algorithmic problems wrapped in a bioinformatic shell, so they might be of interest to a wider audience.

Problem 3 was great: problem sets that are derived from biological data, with partial points awarded to partial solutions, a nice ramp from something that can be solved with pen and paper to something that required a completely automated solution. Also, answers were independent -- the answer to one problem was not the same as the answer to any other problem. The other problems, not so much. Almost everything that was good about problem 3 was bad about problems 1 and 2. A solution to the "tricky" version could be identical to a solution to the "easy" version, and no partial points were awarded despite the solutions going through a series of tests that could have partial points attached to them. The biological basis of problems 1&2 is also very strained. This is especially obvious in the case of problem 2, where the problem designers needed to weasel around with their definition of what constitutes a hydrogen bond, and what constitutes an RNA sequence. I recommend that the problem designers for problem 2 go back and revisit their knowledge of biology, maybe trying out some RNA secondary structure Rosalind problems, such as this one: http://rosalind.info/problems/RNAS/

The last test case of Introns Detection is very interesting.

Finally finish the Qualification Round!