#342 Percolator got worse

[William S Noble]

I just re-ran the performance test, and I noticed that the curve correpsonding to Percolator with tide p-values got noticeably worse. I will attach the two files showing the old results and the new results. As far as I can tell, all the other methods are performing the same, but the cyan "Tide p-value Percolator" line has dropped considerably. We need to track down what change caused this decline.

[Kaipo]

It also looks like the red (Tide p-value q-ranker) and yellow (Tide XCorr Percolator) were affected. I can think of a few things that might have caused this apart from bugs and algorithmic changes: modified parameter defaults, new pin file (refactored xcorr, deltacn, p-value switched to -log10p), new version of Percolator. Not sure yet which is responsible.

[William S Noble]

Seems like the first step would be to back up to previous versions to nail
down exactly which update caused the change, right?
On Oct 6, 2015 12:52 PM, "Kaipo" kaipot@users.sf.net wrote:

It also looks like the red (Tide p-value q-ranker) and yellow (Tide XCorr
Percolator) were affected. I can think of a few things that might have
caused this apart from bugs and algorithmic changes: modified parameter
defaults, new pin file (refactored xcorr, deltacn, p-value switched to
-log10p), new version of Percolator. Not sure yet which is responsible.

---

• [issues:#342] http://sourceforge.net/p/cruxtoolkit/issues/342/
  Percolator got worse*

Status: open
Milestone: post v2.0
Created: Tue Oct 06, 2015 07:15 PM UTC by William S Noble
Last Updated: Tue Oct 06, 2015 07:15 PM UTC
Owner: Kaipo
Attachments:

• stored-performance.png
  https://sourceforge.net/p/cruxtoolkit/issues/342/attachment/stored-performance.png
  (55.9 kB; image/png)
• performance.png
  https://sourceforge.net/p/cruxtoolkit/issues/342/attachment/performance.png
  (55.0 kB; image/png)

I just re-ran the performance test, and I noticed that the curve
correpsonding to Percolator with tide p-values got noticeably worse. I will
attach the two files showing the old results and the new results. As far as
I can tell, all the other methods are performing the same, but the cyan
"Tide p-value Percolator" line has dropped considerably. We need to track
down what change caused this decline.

---

Sent from sourceforge.net because you indicated interest in
https://sourceforge.net/p/cruxtoolkit/issues/342/

To unsubscribe from further messages, please visit
https://sourceforge.net/auth/subscriptions/