Hi Simon,

Yes, the peak with the smallest (in absolute terms) value.

The logic is: we don't know what value was used to threshold the published results. But we can guess that was, at the most, as high as the lowest reported peak (otherwise, the peak would have been thresholded and not reported).

In your example, if you have 4.5 6.7 and 9.8 peaks, the threshold values used could have been, at most, 4.5. Had it been larger (i.e. 4.50001), the 4.5 peak would have been thresholded out and not reported.

I hope this makes sense.

Best,


Anton