<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7638.1">
<TITLE>RE: [Evidence] [Go] question about download page</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<BR>
<BR>

<P><FONT SIZE=2>Hi all,<BR>
<BR>
Yes, I have Karen's list as well as a list of my own that I hope to get through in the near future.<BR>
<BR>
Finding the time is the only problem.&nbsp; Also, I have been hesitant to just make changes as I see fit because perhaps some of these things need broader input.&nbsp; There is a plan to have an ECO content meeting - but at this point I don't know when that would happen.<BR>
<BR>
My current plan is to put together a proposal of changes that I plan to make and send to the evidence list.&nbsp; If I don't hear any objections I'll go ahead and make the changes.&nbsp; If there is disagreement, I'll schedule a skype.<BR>
<BR>
But it will be a few weeks until I get this out - I'm in the middle of moving and barely keeping up with only my most essential work activities.<BR>
<BR>
Sorry for the delays on this.<BR>
<BR>
Michelle<BR>
<BR>
<BR>
<BR>
<BR>
-----Original Message-----<BR>
From: evidence-bounces@genome.stanford.edu on behalf of Karen Christie<BR>
Sent: Thu 4/16/2009 1:26 PM<BR>
To: Midori Harris<BR>
Cc: Chris Mungall; go list; Evidence Code Group<BR>
Subject: Re: [Evidence] [Go] question about download page<BR>
<BR>
I agree that TAS is not placed correctly in the ECO. This is one of the<BR>
things that I commented upon a while ago, when the evidence code<BR>
documentation was under revision. Michelle Gwinn said she's working on the<BR>
ECO now and I recently sent her my whole list of comments/questions on the<BR>
ECO.<BR>
<BR>
-Karen<BR>
<BR>
On Thu, 16 Apr 2009, Midori Harris wrote:<BR>
<BR>
&gt; My guess would be that TAS isn't placed correctly; it's been a last-resort<BR>
&gt; code used in manual annotation, when a curator is inclined to believe<BR>
&gt; something an author says despite being unable to track down an experiment.<BR>
&gt; I'm not aware of it being used for any kind of computational analysis.<BR>
&gt;<BR>
&gt; m<BR>
&gt;<BR>
&gt; On Thu, 16 Apr 2009, Chris Mungall wrote:<BR>
&gt;<BR>
&gt;&gt; We should be using ECO here rather than hardcoding ad-hoc lists of codes in<BR>
&gt;&gt; various pieces of code<BR>
&gt;&gt;<BR>
&gt;&gt; There should be some explicitly declared property in ECO that RCA and IEA<BR>
&gt;&gt; share - either via shared is_a parent or via a relationship to some aspect<BR>
&gt;&gt; of the type of evidence.<BR>
&gt;&gt;<BR>
&gt;&gt; At the moment we have:<BR>
&gt;&gt;<BR>
&gt;&gt; is_a ECO:0000043 ! inferred from in-silico analysis<BR>
&gt;&gt;&nbsp; is_a ECO:00000067 ! inferred from electronic annotation ***&nbsp; [SYNONYM:<BR>
&gt;&gt; &quot;IEA&quot; (related)]<BR>
&gt;&gt;&nbsp; is_a ECO:0000033 ! traceable author statement [SYNONYM: &quot;TAS&quot; (related)]<BR>
&gt;&gt;&nbsp; is_a ECO:0000053 ! inferred from reviewed computational analysis ***<BR>
&gt;&gt; [SYNONYM: &quot;RCA&quot; (related)]<BR>
&gt;&gt;<BR>
&gt;&gt; Which seems dubious. If TAS is placed correctly then we can perhaps use<BR>
&gt;&gt; inferred from in-silico analysis. But I think ECO needs further<BR>
&gt;&gt; re-arrangement or changes such that we have an evidence type &quot;inferred<BR>
&gt;&gt; SOLELY from in-silico analysis&quot;<BR>
&gt;&gt;<BR>
&gt;&gt; On Apr 16, 2009, at 8:01 AM, Valerie Wood wrote:<BR>
&gt;&gt;<BR>
&gt;&gt;&gt;<BR>
&gt;&gt;&gt;<BR>
&gt;&gt;&gt; On the download page, the number of annotations for each organism are<BR>
&gt;&gt;&gt; reported and the number of these which are IEA are specified.<BR>
&gt;&gt;&gt; Should this also specify the number which are non-RCA (as these are<BR>
&gt;&gt;&gt; generally function predictions, with less support than the IEA mappings,<BR>
&gt;&gt;&gt; and users should be made as aware of the 'dangers' of these,&nbsp; as of the<BR>
&gt;&gt;&gt; IEA data).<BR>
&gt;&gt;&gt;<BR>
&gt;&gt;&gt; Val<BR>
&gt;&gt;&gt;<BR>
&gt;&gt;&gt;<BR>
&gt;&gt;&gt; --<BR>
&gt;&gt;&gt; The Wellcome Trust Sanger Institute is operated by Genome Research<BR>
&gt;&gt;&gt; Limited, a charity registered in England with number 1021457 and a company<BR>
&gt;&gt;&gt; registered in England with number 2742969, whose registered office is 215<BR>
&gt;&gt;&gt; Euston Road, London, NW1 2BE.<BR>
&gt;&gt;&gt; _______________________________________________<BR>
&gt;&gt;&gt; Go mailing list<BR>
&gt;&gt;&gt; Go@geneontology.org<BR>
&gt;&gt;&gt; <A HREF="http://fafner.stanford.edu/mailman/listinfo/go">http://fafner.stanford.edu/mailman/listinfo/go</A><BR>
&gt;&gt;&gt;<BR>
&gt;&gt;<BR>
&gt;&gt; _______________________________________________<BR>
&gt;&gt; Go mailing list<BR>
&gt;&gt; Go@geneontology.org<BR>
&gt;&gt; <A HREF="http://fafner.stanford.edu/mailman/listinfo/go">http://fafner.stanford.edu/mailman/listinfo/go</A><BR>
&gt; _______________________________________________<BR>
&gt; Go mailing list<BR>
&gt; Go@geneontology.org<BR>
&gt; <A HREF="http://fafner.stanford.edu/mailman/listinfo/go">http://fafner.stanford.edu/mailman/listinfo/go</A><BR>
&gt;<BR>
_______________________________________________<BR>
Evidence mailing list<BR>
Evidence@geneontology.org<BR>
<A HREF="http://fafner.stanford.edu/mailman/listinfo/evidence">http://fafner.stanford.edu/mailman/listinfo/evidence</A><BR>
<BR>
</FONT>
</P>

</BODY>
</HTML>