Odd parsing of runs associated with a single biosample

[mestato]

Odd behavior found when adding bioproject (PRJEB4873) which has 1 biosample (SAMEA2229347) and 6 experiments each with 1 run (e.g. ERR359850).

After loading biosample XML, Casey found that 7 records were created. One for each run, however, the records do not have db cross references to bioproject or biosample or SRA. And one for the biosample that was not clickable in page view for organism. (see below)

https://hardwoodgenomics.org/organism/Quercus/robur?tripal_pane=gp_Biomaterials

Some of the problems such as cross-references has been manually fixed and we don't have the original XML uploaded by Casey, so first task would be to recreate this and see if the behavior is reproducible.

[CaseyRichards92]

I can try and re load the xml from NCBI and see if this issue replicates itself? Should I delete the current biosamples or will they overwrite ?

[mestato]

If @noahcaldwell7 has a dev version of Tripal with the gene expression module installed, that would be a good testing ground. Since we've already done some manual additions to the biosamples and haven't come up with a good policy on this type of situation, I don't want to overwrite anything or alter the live site yet

[spficklin]

Not sure if this is still a problem but I tested loading the biosample mentioned here using the updated loader in PR #392 and it only created one BioSample record. So perhaps it's fixed?