Comments for Proteomics.ME

Comment on UNC TIM login by JO

JO — Tue, 05 Jul 2011 17:20:40 +0000

I can’t even tell you how many times I’ve used your page/link to get to TIM! Most helpful page ever.

Comment on UNC TIM login by EJM

EJM — Tue, 28 Jun 2011 16:17:57 +0000

Thank you so much! I’ve been through the “where in the world is the TIM link” on UNC’s website and it was soooo frustrating.

Comment on UNC TIM login by H

Fri, 04 Mar 2011 22:38:06 +0000

Thanks so much! This was a big help today when I needed to approve my timesheet from home.

Comment on UNC TIM login by C

Mon, 08 Nov 2010 18:01:00 +0000

Thanks for the link. Service to all. You are awesome.

Comment on TandemFit animation by Jainab Khatun

Jainab Khatun — Fri, 20 Aug 2010 14:51:37 +0000

Great!
If you can level the ion type while traveling.

Comment on How to Fake a Great E Value by Jainab Khatun

Jainab Khatun — Fri, 13 Aug 2010 19:01:59 +0000

Brain, Morgan or whoever has concerned with E_values,
Well, may be there is no standard for E-Values, such as Brian pointed out that we can change the values of E_Values simply by squaring the scores. But in my understanding the relative values does not change, just you get different values. For examples in Brian’s first scoring method the difference in E-Values is 10^7. Now that does not mean that the identification for spectrum A is 10,000,000 times more confident than the spectrum B unless using the same scoring method and sufficient amount of standard data someone actually verifies that. 10^7 difference may actually mean 10% increase in confidence and we have to find out using standard data. Similarly for the second scoring method even though the difference in E-Values is 10^34, that may mean only 10% increase in confidence and in this case E-Values of 10^-20 may actually mean false positive. Therefore in my understanding E-Values are relative and we have to find our own standard, I mean we have to find out does E-Value of 10^-6 actually mean 1 in a million FP. I do trust E-values if you define your standard using sufficient amount of data.

Comment on proteomic database fields by Peter

Peter — Mon, 21 Dec 2009 23:22:16 +0000

I have put together a normalized schema for this, but will need help with the specifics. Currently I do not store information about number of relations to a peptide/sequence because this can be computed efficiently.

Using a database like this may make implementation under condor or other parallel computing frameworks more efficient, since sequences may be demand loaded.

I would post a drawing here but I do not know how this tool accommodates drawings…

Comment on Objective-C MySQL API by nollkoll

nollkoll — Thu, 22 Oct 2009 19:53:09 +0000

salve,

same here. it’s strange there is next to nothing going on for mysql on os x. ok there is sqlite but that is clearly not the same thing. i dont know much about core data framework but ideally it would be nice if it worked with mysql too as persistence framework…

Comment on Tandem Mass data hash by Brian

Brian — Wed, 14 Oct 2009 15:43:28 +0000

Of course: I should keep the bins relative to the max values for a specific spectrum. This makes perfect sense as we would only be comparing spectra with similar precursor masses. Also, it is obvious the method I have chosen for parceling along the y-axis is not the best.

Comment on Tandem Mass data hash by Morgan

Morgan — Wed, 14 Oct 2009 14:56:51 +0000

Jainab, that is a good suggestion. But saying “so that we can be sure that this won’t work” is a bit of a misnomer. I don’t think these experiments proved anything of the sort that the method won’t work. In fact, contrary to what Brian wrote, I find these first experiments quite encouraging, given that they are exactly that: _first_ experiments.