When a user clicks on a paired-analysis row we want to return samples with at least those analyses (i.e. there could be more). Since that upset plot filter only filters samples and not analyses, we currently keep the MG-MT-NOM row as well when we click on the MG-NOM row and don't hide the fact that there are some with MT. So those 43 MT are from samples that match the user search (have MG and NOM).
See March 2023 comments from https://github.com/microbiomedata/nmdc-server/issues/1072
When a user clicks on a paired-analysis row we want to return samples with at least those analyses (i.e. there could be more). Since that upset plot filter only filters samples and not analyses, we currently keep the MG-MT-NOM row as well when we click on the MG-NOM row and don't hide the fact that there are some with MT. So those 43 MT are from samples that match the user search (have MG and NOM).
https://private-user-images.githubusercontent.com/81305/309838596-e8c386c5-b21e-4fb0-99e4-1f45c7ce0d2f.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3MTA0NTY2MDEsIm5iZiI6MTcxMDQ1NjMwMSwicGF0aCI6Ii84MTMwNS8zMDk4Mzg1OTYtZThjMzg2YzUtYjIxZS00ZmIwLTk5ZTQtMWY0NWM3Y2UwZDJmLnBuZz9YLUFtei1BbGdvcml0aG09QVdTNC1ITUFDLVNIQTI1NiZYLUFtei1DcmVkZW50aWFsPUFLSUFWQ09EWUxTQTUzUFFLNFpBJTJGMjAyNDAzMTQlMkZ1cy1lYXN0LTElMkZzMyUyRmF3czRfcmVxdWVzdCZYLUFtei1EYXRlPTIwMjQwMzE0VDIyNDUwMVomWC1BbXotRXhwaXJlcz0zMDAmWC1BbXotU2lnbmF0dXJlPWQwZGQ4NTdkZjNiZjI2MDhiZjEyNTJkOGVkYTU2OWQzYjI3Y2JkYmY3NDAzZDY1YjJmNGE0YzVhMDk2MGMzY2EmWC1BbXotU2lnbmVkSGVhZGVycz1ob3N0JmFjdG9yX2lkPTAma2V5X2lkPTAmcmVwb19pZD0wIn0.DFR8Qm8WwwJEpAn8hdNvI2dRA112MMl_Hm-gX_r9NZ4
determine if clicking on MG-NOM (134) should only return those records or should be the minimum filtering criteria.