two new problems discovered

[Bogdan]

hi all,

two problems were discovered since yesterday with the PWM finder:

1. Submitting a task with a fractional matrix which has values equal
to 0.0 produces empty results file.
2. Segmentation fault occurs for the unknown reason in PWM finder on
some legitimate integer frequency matrices.

Both problems are being looked into. At the moment, there are two
users affected with these two problems.

[Bogdan]

Problem #2 is now fixed.

It occurred when user's email address was longer than I initially
expected.

[Bogdan]

Problem #1 is now fixed.

> 1. Submitting a task with a fractional matrix which has values equal
> to 0.0 produces empty results file.

The problem was with -Inf values assigned to PWM cells, which
correspond to 0-value cells of the user-submitted fractional PFM.


To (technically) avoid -Inf values in the PWM matrix, I've applied a
small and fixed pseudocount (10E-9) to all searches performed with
fractional matrices.


Effectively, zero-value positions of the user-supplied fractional
matrix translate to PWM value -29.897, which is ~10 times less than
the lowest (non-zero) PWM values.