When trying to load a CSV file of runs obtained from the NCBI Run Selector using the Biosample loader I do not see any elements in the 'CVTerm Field Configuration' section. I don't get an error telling me anything is wrong either. It just doesn't do anything. Below is a screeshot of what I see.
I've tracked down the problem being the following lines of code
The problem is that "sample_name" is hardcoded as the column that should contain the biosample name. In the CSV file I obtained from NCBI the column is titled 'Sample Name' (capitalized and a space between it).
I will submit a pull request to fix this. But I think a good way to handle this would be to not hardcode a column name and let the user specify the name of the column that has the biosample ID. Otherwise it will always be a cat and mouse game of keeping up with how NCBI changes file formats. My quick PR fix will just be an updated regular expression to find "Sample Name".
spficklin
commented
2 years ago
When trying to load a CSV file of runs obtained from the NCBI Run Selector using the Biosample loader I do not see any elements in the 'CVTerm Field Configuration' section. I don't get an error telling me anything is wrong either. It just doesn't do anything. Below is a screeshot of what I see.
I've tracked down the problem being the following lines of code
https://github.com/tripal/tripal_analysis_expression/blob/35a51746462a81b9a046aa81d8992d285961b1ee/tripal_biomaterial/includes/TripalImporter/tripal_biomaterial_loader_v3.inc#L597-L603
and here:
https://github.com/tripal/tripal_analysis_expression/blob/35a51746462a81b9a046aa81d8992d285961b1ee/tripal_biomaterial/includes/TripalImporter/tripal_biomaterial_loader_v3.inc#L619-L625
The problem is that "sample_name" is hardcoded as the column that should contain the biosample name. In the CSV file I obtained from NCBI the column is titled 'Sample Name' (capitalized and a space between it).
I will submit a pull request to fix this. But I think a good way to handle this would be to not hardcode a column name and let the user specify the name of the column that has the biosample ID. Otherwise it will always be a cat and mouse game of keeping up with how NCBI changes file formats. My quick PR fix will just be an updated regular expression to find "Sample Name".