On 2016 Jun 29, Bill Noble commented:

Hi Zhibin, thank you for your comment. Yes, as we describe on p. 1148 in the paragraph beginning "Our primary goal in this article ...", there are many methods described in the literature for obtaining approximately calibrated scores. These are variously referred to as E-values or p-values, though the precise semantics of any such score depends on the method used to generate it. One of the main points of our article is to introduce an independent method to help evaluate the calibration properties of these types of scores. For example, Figure 3 shows the calibration properties of the MS-GF+ E-value. --Bill Noble & Uri Keich

On 2016 Jun 24, Zhibin Ning commented:

Actually, quite a few search engine and downstream workflows do not use "score" to rank the peptide, instead, p-value is more appropriate to do the "second job" claimed in this manuscript. P value (or E value), in some sense, is the calibrated "score".

On 2016 Jun 24, Zhibin Ning commented:

Actually, quite a few search engine and downstream workflows do not use "score" to rank the peptide, instead, p-value is more appropriate to do the "second job" claimed in this manuscript. P value (or E value), in some sense, is the calibrated "score".

On 2016 Jun 29, Bill Noble commented:

Hi Zhibin, thank you for your comment. Yes, as we describe on p. 1148 in the paragraph beginning "Our primary goal in this article ...", there are many methods described in the literature for obtaining approximately calibrated scores. These are variously referred to as E-values or p-values, though the precise semantics of any such score depends on the method used to generate it. One of the main points of our article is to introduce an independent method to help evaluate the calibration properties of these types of scores. For example, Figure 3 shows the calibration properties of the MS-GF+ E-value. --Bill Noble & Uri Keich