7

We performed some bioinformatic work during thesis. Output of these analysis are more than 10 tables with 2000 rows each. We cannot add these in the printed version of the thesis, so we thought about:

  • link to the online version, since paper containing this data was published.

  • give a CD-ROM with the thesis, but for me this physical format is obsolete.

Maybe you have a better alternative or agree with the first one like me.

8
  • 4
    For online version, you may want to drop it into a well-known data repository or even the arXiv. (Please don't just put the online version on your group website, unless you intend to permanently maintain a stable URL and provide archival storage of the information.) Jun 29, 2015 at 8:41
  • @Willie Wong, do you think giving the link to the journal where paper was published is not adequate? I mean direct link to data.
    – biotech
    Jun 29, 2015 at 9:30
  • Can you please clarify: 1) What sort of thesis is this; in particular, is it a published thesis? 2) Who is we?
    – Wrzlprmft
    Jun 29, 2015 at 9:42
  • It is not clear to me what you meant by "Paper containing this data was published". Do you mean that the data is included in a (long) paper published somewhere? Do you mean that the data is included as online supplementary information for a paper somewhere? Or do you mean simply that a paper based on that data is published somewhere? In the first two cases then yes, you can just link to the published version. But in case two depending how your journal handles supplementary information it may still be a good idea to publish the data elsewhere in addition. Jun 29, 2015 at 10:53
  • @Willie Wong, the data was included as online supplementary information for a paper.
    – biotech
    Jun 29, 2015 at 10:58

2 Answers 2

12

Output of these analysis are more than 10 tables with 2000 rows each. We cannot add these in the printed version of the thesis ...

Even if you could (say for instance you hate trees and you don't pay for your thesis), do you think anybody would want to have that kind of data in physical format?

Both alternatives are valid, but as you said optical media is becoming more and more obsolete. If you really want to provide the data in a physical form, you can distribute it as a usb-stick. As the capacity of sticks increase, smaller drives become dirt cheap. You should be able to get 1GB, 2GB or even 4GB for relatively small price in hundreds.

The other option, online availability, is the better alternative in general. However the maintenance of data is important as others have mentioned. Both in terms of availability and proper cataloging try to use well known and established repositories.

What I don't understand however, is why you feel the need to "republish" findings, if you have already deposited this data together with a prior publication. Just refer to the data as "Table/Supplementary Info X from Paper Y"

3
  • 1
    It's also possible that physical media could be a requirement of the university.
    – MJeffryes
    Jun 29, 2015 at 10:58
  • I will try online availability option, physical media storages are obsolete, even USB sticks.
    – biotech
    Jun 29, 2015 at 11:00
  • 5
    If you want it to fit well in the printed copies, you can buy SD cards :-)
    – yo'
    Jun 29, 2015 at 11:33
5

While publishing data sets on arXiv might be a decent option, I suggest you to consider posting your research results and artifacts to figshare or Zenodo. Both services allow artifacts to be not only citable, but also discoverable (via DOI assignment). Version control is supported as well, but the advantage of Zenodo in this regard in comparison with figshare is Git and GitHub integration.

4
  • 1
    It sounds like there is already a DOI available, since the asker has said that they have published a paper with these data. Is it necessary to upload somewhere else too? Is figshare or Zenodo more likely to stick around than the journal?
    – MJeffryes
    Jun 29, 2015 at 10:27
  • 6
    @MJeffryes: The beauty of the repositories that I recommended is that both allow to create citable artifacts, which, in this case, would refer to citable data sets (with separate DOIs). This is a very popular trend today in the light of overall bigger trend of wider promotion and use of the reproducible research paradigm. As for the longevity, I am pretty confident in both services, especially Zenodo. IMHO it will stick around, at least, as long as a typical journal, if not longer, since it is EU-funded and backed by more than solid CERN IT infrastructure. Jun 29, 2015 at 10:51
  • 1
    It sounds like Zenodo is the perfect solution for this problem. I will certainly bookmark it for possible future use...
    – Floris
    Jun 29, 2015 at 17:19
  • 1
    @Floris: You are welcome. I am glad to be able to help. Jun 29, 2015 at 21:53

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .