glycinemine-5103 has an orphaned GWAS trait

[sammyjava]

glycinemine-5103=> select * from trait where name='Branching' and gwasid is not null;
 description |    id     |   name    |         primaryidentifier         | qtlstudyid |  gwasid   | datasetid |             class             
-------------+-----------+-----------+-----------------------------------+------------+-----------+-----------+-------------------------------
             | 298000489 | Branching | mixed.gwas.Fang_Ma_2017:Branching |            | 298000034 | 298000001 | org.intermine.model.bio.Trait

glycinemine-5103=> select * from gwasresult where traitid =298000489;
 id | pvalue | primaryidentifier | markername | gwasid | traitid | datasetid | class 
----+--------+-------------------+------------+--------+---------+-----------+-------
(0 rows)

[sammyjava]

What's REALLY WEIRD is that GWAS doesn't have any Branching trait!!! So the orphan status is correct, it's not in the results file, but the trait mixed.gwas.Fang_Ma_2017:Branching should not exist in the mine!!!

[sammyjava]

The obo file in that collection includes Branching, but there is no Branching trait in the results file. It should have failed validation. So fix the GWASCollectionValidator to be sure that the trait names in the obo file match those in the results file, going both ways.

[sammyjava]

Done. And there was another one, Seed palmitic.

## Validating glyma collection mixed.gwas.Fang_Ma_2017
 - glyma.mixed.gwas.Fang_Ma_2017.obo.tsv.gz
 - glyma.mixed.gwas.Fang_Ma_2017.result.tsv.gz
## INVALID: OBO file glyma.mixed.gwas.Fang_Ma_2017.obo.tsv.gz contains trait Branching that is not present in result file.
## INVALID: OBO file glyma.mixed.gwas.Fang_Ma_2017.obo.tsv.gz contains trait Seed palmitic that is not present in result file.

[sammyjava]

So, it turned out that only two GWAS collections had errant traits, both in the OBO file. Li_Zhao_2019 had Seed oil and Seed protein that were not in the result file, so I'll yank those as well.