the last step for calculate index of result---metagenomic gene prediction

Question

[jinxyhucy521]

Dear guys! I have a problem before obtaining the final result.I do some research about metagenomic gene prediction.I have predicted the testing data in my own model.The predicting results have the coding ORFs and non-coding ORFs.And I also have the lables of these ORFs befores testing!I want to know how to calculate the Sn and Sp.Should it be based the number of ORF or the number of base pairs? I have another problem.The ORFs are abstracted from the fragments.some of them are overlapped,and how i shoule deal with this problem? I hope I can get useful information about the above two problems.Thanks a lot!!

[Giovanni M Dall'Olio]

Could you modify the title of this post, putting something more specific to your query? Imagine how many people use biostar for the last step of their paper, it would be a mess if everybody used the same title.

Answer 1

[jinxyhucy521]

Thanks for your remind,and I have changed it by myself!!