🐛 Found instance of HuBMAP dataset in wrong as-cell-summary

[andreasbueckle]

@bherr2
The fig. below shows as-cell-summaries for two AS in the female heart, annotated with Azimuth:

The datasets contributing to these are here:
https://github.com/x-atlas-consortia/hra-pop/blob/main/output-data/v0.10.2/atlas-as-cell-summaries.jsonld#L11329-L11343

I did some deep digging tonight to build a bar graph showing CTs per dataset that contribute to the AS shown in the fig. below. However, if you look at the provenance for the first dataset (https://portal.hubmapconsortium.org/browse/dataset/d6e6c8e452ed628425e9e928306a6db0#provenance), you will notice that the donor is male while the dataset is listed in the aggregated_summaries for the female, Azimuth annotation.

I only noticed this discrepancy, because I was getting cell summaries for the datasets in the aggregated_summaries from this report: https://github.com/x-atlas-consortia/hra-pop/blob/main/output-data/v0.10.2/reports/atlas-ad-hoc/datasets-ct-bm-data.csv
But I only saw male summaries for this particular dataset (https://entity.api.hubmapconsortium.org/ancestors/d6e6c8e452ed628425e9e928306a6db0).

My code to build the bar graphs is here: https://github.com/cns-iu/hra-construction-usage-supporting-information/blob/main/hra_pop/figure_1_g_hra_pop.ipynb

Can we investigate?

[andreasbueckle]

PS: I found the issue:
This dataset is from a male donor but was registered into the female heart: https://portal.hubmapconsortium.org/browse/sample/60da3463ccefe031ad887014140a341e.json

\"target\": \"http://purl.org/ccf/latest/ccf.owl#VHFHeart\"

This dataset is also listed in this SPARQL query, which returns a list of all tissue blocks where the donor sex does not match the sex of the reference organ in the extraction site/RUI location: https://grlc.io/api-git/hubmapconsortium/ccf-grlc/subdir/ccf/#/default/get_sex_mismatches

Need Dan Q to resolve so this is fixed in hra-pop v0.10.4.