> <bigger>1. From the corrected values for the regions selected for
> calibration, it seems to me that the FPmax for HS2 is 118, not 138, and
> for combined it is 518, not 538.  Am I correct, or did I subtract
> incorrectly?

The region selected for the calibration on HS2 was actually 11240-11510
(and not 11240-11490, as I wrote by mistake).  This explains the 20 bp
in the difference. I am really sorry about that. 

> 2.  From section 2.0 of Results, a paragraph says:
> 
> 
> For example, the results of the optimization for
> <italic>infocon</italic>'s anchor value are shown in Fig. 3.  The
> regions encompassing HS2, HS3 and the <italic>HBB</italic> promoter
> were examined by increasing the anchor value in increments of 0.001
> from 0 to 2, holding the minimum length constant at the best value for
> a particular region. The resulting columns for each anchor value were
> compared to the reference set of known functional sequences (see
> below).  A clear minimum cost can be seen at a certain anchor value for
> each of the three regions.  Interestingly, a slightly different anchor
> value, and a different minimal cost, is obtained for each region."
> 
> 
> The Methods says the range was -10 to 2; was this a different
> experiment?]

No: the experiment was run for the range -10 to 2, however the results
(#fp, #fn) were constant on the range -10 to 0. I believed that 0-2 was 
actually the informative range, because this is where we observe the 
variation. We should probably use the same values in both sections of 
the paper, for consistency.  


> 3. Under section 2.2 in Methods, we give a numerical example for
> infocon.  Let's provide the answer as well, such as :
> 
> 
> "The resulting value is _____ for the information content."

" The resulting value is 2.0884482 for the information content."

>Maybe more to come,
>
>Thanks,
>
>Ross


Thank you!
Liliana