[XMLPipeDB-developer] TallyEngine bugs

[Kam D. <kda...@lm...>]

Hi,

I'm starting this as a new thread (hopefully) so we can keep the threads 
separate by task.

There are issues with the TallyEngine for each species, maybe they are 
related, but I don't know; I'm linking to their testing report pages where 
you can look at the results screen shot and the OriginalRowCounts data 
directly.

Kam

New species:
Mycobacterium tuberculosis 
https://sourceforge.net/apps/mediawiki/xmlpipedb/index.php?title=Gene_Database_Testing_Report_M._tuberculosis_20090817 
and P. aerugenosa 
https://sourceforge.net/apps/mediawiki/xmlpipedb/index.php?title=Gene_Database_Testing_Report_P._aerugenosa_20090817
* OrderedLocusNames is not being tallied
* UniGene is not being tallied (UniGene doesn't exist for these two, but in 
other TallyEngine results for other species, it is actually listed with 
counts of zero.  We should be consistent about counting it.)

Old species:
E. coli K12 
https://sourceforge.net/apps/mediawiki/xmlpipedb/index.php?title=Gene_Database_Testing_Report_E._coli_K12_20090819
* There are discrepancies between the TallyEngine results and the gdb
** EchoBASE has 4022 records in gdb, not 4028
** EcoGene has 4190 records in gdb, not 4199
** Blattner has 4328 records in gdb, not 8352
* W3110 is not being counted (or is being conflated with Blattner)
* EchoBASE, EcoGene, and Blattner are appearing twice in results with 
second instance giving -1 result for XML Count.
* UniGene is not being tallied (UniGene doesn't exist for these two, but in 
other TallyEngine results for other species, it is actually listed with 
counts of zero.  We should be consistent about counting it.)

Arabidopsis 
https://sourceforge.net/apps/mediawiki/xmlpipedb/index.php?title=Gene_Database_Testing_Report_Arabidopsis_20090819
* Discrepancy between TallyEngine and OriginalRowCounts for the following 
systems:
** GeneId
** RefSeq
** TAIR
** UniGene
** In these cases, gdb has fewer records.
* TAIR and UniGene are repeated in TallyEngine results, showing a -1 for 
the XML Count for the second result.

Plasmodium 
https://sourceforge.net/apps/mediawiki/xmlpipedb/index.php?title=Gene_Database_Testing_Report_P._falciparum_20090819
* Discrepancies in TallyEngine Results versus OriginalRowCounts
** GeneId TallyEngine has 5264, but OriginalRowCounts is 5260
** OrderedLocusNames in gdb is 5336 and much less in TallyEngine, also XML 
Count and Database Count are off by 1.
* OrderedLocus and UniGene repeated twice in results with a -1 given for 
the XML Count the second time.

Vibrio 
https://sourceforge.net/apps/mediawiki/xmlpipedb/index.php?title=Gene_Database_Testing_Report_V._cholerae_20090820
* OrderedLocusNames counts are off by one between the XML and database 
(slash issue?); also, because of the underscore issue, remember that all of 
these IDs got duplicated in the gdb +/- the underscore.
* OrderedLocusNames and UniGene appear twice in results, second time with 
-1 in XML Count.