deltaCn

[Bill Noble]

This email was not sent to the Google group, but I'm replying here in case others have this question:

Dear Prof. William Stafford Noble

This is Naiping Dong, a doctoral student in Central South University

in China. When I used Crux to learn the XCorr scores, I found that in

some results, the deltaCn scores would be as large as more than 100.

Since in my knowledge, this score is the difference between normalized

XCorr scores of the first and second rank. What happened in this

calculation in Crux?

It would be very appreciated to recieve your reply.

Thanks very much.

Yours Sincerely

Naiping Dong

2012/2/9

The answer is that deltaCn is the difference between the top-ranked score and the given score, divided by the top-ranked score.  If the top-ranked score is very close to zero, then the deltaCn could become quite large.  If you have a specific example for which you think the deltaCn is being computed erroneously, please send it along and we will take a look.

Thanks.

Bill