Some submissions have undergone a curation process at GEO, giving them a GDS accession. NCBI has not been curating data for several years, so the GDS accessions are historical and now represent a pretty small proportion of the GEO information. A GDS and GSE represent pretty much the same type of data, a collection of samples. A GSM is a single sample in GEO. Therefore, GDS and GSE contain one or more GSMs.
I have already gone though the GEO database, but it doesn't answer my question. In GEOmetadb, I found some unique GDS IDs like GDS15, GDS16, GDS17, GDS18, GDS19, GDS20, GDS21, GDS30 and GDS31 with common GSE ID - GSE18. My question is on what basis are these GDS IDs are common to GSE18. Also, they belong to different GPL IDs like GPL51, GPL52 and GPL53.
See also the GEO FAQ.
Hi. Sean Davis,
I have already gone though the GEO database, but it doesn't answer my question. In GEOmetadb, I found some unique GDS IDs like GDS15, GDS16, GDS17, GDS18, GDS19, GDS20, GDS21, GDS30 and GDS31 with common GSE ID - GSE18. My question is on what basis are these GDS IDs are common to GSE18. Also, they belong to different GPL IDs like GPL51, GPL52 and GPL53.
Please read the link in my comment, this is described there.
GSE18 has nine different platforms (microarray types). The GDS records each represent one of those microarray types.