[Bioc-devel] hgu95aprobe missing sequences

[Marc Carlson]

Seth Falcon wrote:
> "James MacDonald" <jmacdon at med.umich.edu> writes:
>   
>> 3.) Both the yeast2 and ygs98 annotation packages appear to be
>> busted. In both cases there appear to be far fewer probeset keys in
>> the envs than are reported. For instance, for the ygs98 package:
>>
>>     
>>> length(ygs98CHR)
>>>       
>> [1] 6776 ygs98()
>>     
>
> Marc has identified a bug in the yeast2 AnnBuilder scripts and is
> working on finding the root cause and a fix.  We will at least be able
> to resolve this for the next release if not sooner.
>
>
> + seth
So I have spent some time on this problem and come to the following
conclusions:
1) there is more than one bug within AnnBuilder as pertains to the
Affymetrix Yeast packages.
2) these bugs seems to have been around for a long, long time.  I don't
know if these packages have ever had all the IDs mapped/available for yeast.
3) I have found one of these bugs and corrected it which places most of
the missing IDs into the yeast2 package, but unfortunately this
particular fix will do almost nothing for the ygs98 package as it has
another as of yet unidentified issue.

I have therefore checked in a fix for of AnnBuilder, and made a new
yeast2 package which I will push out next week when I can get Herve to
help me.  Since these packages have been around for a long time, I
intend to focus more of my future time on rapidly making the new sqlite
based versions of these packages ready for the next release.  The
environment styled packages should be able to be generated from these
sqlite packages as well which I am hoping will allow us to make a clean
break from this.  So from the users perspective, this should hopefully
mean that the yeast packages should just mysteriously look better with
the next release.