Dear all,
I have an extremely long list of unannotated CEL files and I would like to fetch metadata from GEO in bulk. Any advice for that?
Thanks in advance!
•
written
by
lessismore
You can use the Entrez utilities to get the information you need:
$ esearch -db gds -query GSM85508 | efetch
1. Basal-like breast cancer tumors
Analysis of sporadic basal-like cancer (BLC), BRCA-associated breast cancer, and non-BLC tumors. Sporadic BLC are phenotypically similar to BRCA1-associated cancers. Results provide insight into the molecular pathogenesis of BLC and BRCA1-associated breast cancer.
Organism: Homo sapiens
Type: Expression profiling by array, transformed count, 4 disease state sets
Platform: GPL570 Series: GSE3744 47 Samples
FTP download: GEO (CEL) ftp://ftp.ncbi.nlm.nih.gov/geo/datasets/GDS2nnn/GDS2250/
DataSet Accession: GDS2250 ID: 2250
2. Human breast tumor expression
(Submitter supplied) Gene expression for 47 human breast tumor cases; (* normalized by GCRMA for global expression analysis) Keywords: Type
Organism: Homo sapiens
Type: Expression profiling by array
Dataset: GDS2250 Platform: GPL570 47 Samples
FTP download: GEO (CEL) ftp://ftp.ncbi.nlm.nih.gov/geo/series/GSE3nnn/GSE3744/
Series Accession: GSE3744 ID: 200003744
3. [HG-U133_Plus_2] Affymetrix Human Genome U133 Plus 2.0 Array
(Submitter supplied) Affymetrix submissions are typically submitted to GEO using the GEOarchive method described at http://www.ncbi.nlm.nih.gov/projects/geo/info/geo_affy.html June 03, 2009: annotation table updated with netaffx build 28 June 06, 2012: annotation table updated with netaffx build 32 June 23, 2016: annotation table updated with netaffx build 35 Protocol: see manufacturer's web site Complete coverage of the Human Genome U133 Set plus 6,500 additional genes for analysis of over 47,000 transcripts All probe sets represented on the GeneChip Human Genome U133 Set are identically replicated on the GeneChip Human Genome U133 Plus 2.0 Array. more...
Organism: Homo sapiens
602 DataSets 4699 Series 58 Related Platforms 133119 Samples
FTP download: GEO ftp://ftp.ncbi.nlm.nih.gov/geo/platforms/GPLnnn/GPL570/
Platform Accession: GPL570 ID: 100000570
4. T92 U133p2
Organism: Homo sapiens
Source name: T92
Platform: GPL570 Series: GSE3744 Dataset: GDS2250
FTP download: GEO (CEL) ftp://ftp.ncbi.nlm.nih.gov/geo/samples/GSM85nnn/GSM85508/
Sample Accession: GSM85508 ID: 300085508
To do a batch search you'll need to write a script to read each of your GSM accessions and send a query to NCBI with the above command.
Thanks very useful. If for each GSM i would like to retrieve only the GSE? Could you comment on this?
I guess you could just use the grep command line tool to retrieve the GSE numbers:
grep "Series:"
I believe GEOmetadb could help you:
https://gbnci-abcc.ncifcrf.gov/geo/gsm.php R package: http://bioconductor.org/packages/2.2/bioc/html/GEOmetadb.html
Thanks, what im searching is problably the esummary tool, but i tried for 2 ids in sra database and it doesnt work.
esummary.fcgi?db=database&id=uid1,uid2,uid3,...
Log in to answer this question.
Could you post the GSE ID. I'll try and get back to you.
Dont have the GSE, just have the GSM IDs
Could you please give me a couple of those?
Hey, let's work with GSM85508 for example.
Open https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM85508 and scroll down. You will find a label named "Series", beside which the GSE ID is mentioned. Click on the GSE ID, and at the bottom of the page, there will be a link to a compressed file. The compressed file would contain the .cel files
Thanks but im interested in the bulk metadata download. should be with this but its not working:
Solved: It's quite tricky
e.g. https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi?db=sra&id=010170,..
First you extract the experiment ID associated to the GSM ID, then you can redo the search to extract the values for the specific term you are interested. If anyone has a faster solution it would be very welcome
Check out: https://poe.mm.di.uoa.gr/
Don't know about other people, but I would never use a website that has this type of a front page. Without any explanation about anything, are we supposed to be convinced that a "Welcome to Public Omics Explorer" title is something worth exploring? If you are in charge of that website, I suggest you seriously reconsider its front page.
Agreed. Going to move this "answer" to a comment (for now), since we have no idea if the site/tool is relevant to the original question.
If you are requiring "registration" for access, it is only fair that you describe what people are signing up for, including functionality/conditions/limits up front. If this is a limited, partially free and/or a commercial tool that should also be disclosed.
Thanks for the feedback that’s totally fair.
Just to clarify, you can log in as a guest (no registration needed) and explore the full functionality of the app. There’s also detailed documentation within the tool, along with the associated peer-reviewed paper (DOI: 10.1016/j.csbj.2025.11.004) that explains the methodology and use cases.
I appreciate the comments about the front page we’ll definitely work on making the purpose and value clearer upfront.