I would like to depend on Biosample database for sample metadata rather than SRA. Do all SRA samples have a corresponding sample in Biosample? I've browsed a handful of studies, and found that they do. However, I'm wondering if anyone knows better than I.
2 answers
According to the rules of NCBI, it is mandatory for each SRA submission to have a registered corresponding BioSample: "Registering project and biological samples at the NCBI BioProject and the BioSample databases is a prerequisite for any public SRA submission" (https://www.ncbi.nlm.nih.gov/sra/docs/submitbio/).
I have accessed many public datasets and all of them have a BioSample ID, an experiment ID and run ID(s). Many but not all have a secondary sample ID which links to GEO.
Log in to answer this question.
I know that samples submitted to NCBI are regularly pulled into BioSamples by Ensembl. I'm not sure if that happens in the other direction.
Check this directory for SRA reports. I would look at the datalist directory to see if those files help.