I had a quick look online. There seems to exist many different website of database archiving. Some data might be free of charge while some others might not be. I found things such as Dryad, TreeBase, DataOne or GeneBank. I think that several journals (American Naturalist, Molecular Ecology, Science and Nature for examples) force their authors to make their data available

General Question

Can you make a short review of what different database exist, what are their properties (what kind of data, what license and fees) and how to seek for data (are datasets associated to keywords or do we have to look for a paper and then look for their data).

Case study

Imagine for example, you are looking for SNP data of a spatially explicit population that recently underwent a range expansion (might be too accurate, you may want to drop down some specifications). How would you go about to look for this kind of data? (No need to pull out data but only to describe a procedure).

  • 1
    $\begingroup$ Please leave a comment when downvoting $\endgroup$ – Remi.b May 26 '15 at 15:04
  • 1
    $\begingroup$ I am working a lot with databases, and I have to say your question is too broad. There are heaps of different databases and thousands of ways to use them. Even describing GenBank in detail would make half a book. In you case, if you can describe the biological question so well, try looking for relevant publications, take the database identifiers they give for the data they used and look at what you get. $\endgroup$ – skymningen May 27 '15 at 11:47
  • $\begingroup$ @skymningen Thanks for your comment. So, for the case study I am asking you would not try to look with keyword in a specific database but you would rather look for articles that may have potentially uploaded their datasets? Seems like it would take a long time to find relevant data. Would it be too broad to ask for a list of the different databases that exist along with their very basic feature such as broadly speaking, the nature of the data (genetic/phenotypic for example) or the field that have produced those data (ecology/biochemistry/evolutionary biology/...)? $\endgroup$ – Remi.b May 27 '15 at 14:18
  • $\begingroup$ That is still too many databases and I by far do not know all of them. I am specialised in -omics data, so I could mainly guide you there. The NCBI-databases are your best guess to get datasets, but the descriptions given there are often hard to understand if you did not read the publication. It is much easier to skim over a publication and decide if the data they used or produced is something that could help you. To help you with your case study I am already failing at the meaning of "spatially explicit", sorry. $\endgroup$ – skymningen May 28 '15 at 6:07
  • $\begingroup$ @skymningen Ok. Well I guess you actually made the answer one could give through your two comments. Could make an answer out of your comments? Maybe something along the lines of: 1; One cannot summarize all data bases, there are too many of them 2; Because the datasets usually don't have enough metainformation (and eventually because tagging is not super efficient), your best bet is to look at the literature directly. $\endgroup$ – Remi.b May 28 '15 at 13:05

As a bioinformatician, one of my tasks is quite often to get information from databases or find the correct database to compare to. There are several big databases around, but even those are way too many for any single person to know. There are databases specialising on genetic data, some specialising on taxonomic information, some on ecological information and then there are species specific databases which collect all kinds of data from a given species, genus or clade.

For me the biggest source of generalised information would be the NCBI databases, including GenBank for nucleotide sequence, their Protein database and maybe dbSNP for, well, SNPs. Then there are multiple 16S databases I have used, there is UCSC for the basic data about the human genome, flyBase for Drosophila,... I think the point should be clear. And those are only the ones that just sprang to my mind.

If you don't have a very exact question to answer like "Is there a gene similar to gene x from Drosophila present in any Bacteria?" it can be very hard to both choose the correct database and also find the important information in it efficiently.

Then there is the problem of metadata. I might be able to find sequencing data from the gut microbiome of a human treated with antibiotics quite easily. If I am looking for a specific antibiotic given over a minimal time of seven days to a different healthy humans, preferably including at least one breastfed child and it's mother (which was the one taking the antibiotics, not the child), then it's getting complicated. Here it is generally way easier to scan PubMed for some relevant publications on the matter and find out if they used published datasets or published their own.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.