Saving Experimental Files

From Heit Lab Wiki
Jump to navigation Jump to search

Logging Computer Files

It is not uncommon for experiments to generate large quantities of computer files containing raw data, analyzed data, statistics, intermediary stages of analysis and publication-quality images. It is critical these files be saved in a fashion whereby they can be easily retrieved and assigned to the specific experiment in the laboratory notebook which generated the files. Because of this it is critical that you save your files using the following guidelines:

  1. All files for all laboratory users should be saved on the laboratory file server, which is backed up nightly. A personal computer must never be the sole site where data is stored.
  2. CD’s, DVD’s, blu-ray disks, memory sticks and most other forms of removable media are not sufficient for primary data storage or backup due to the risk of damage or loss.
  3. Raw and analyzed data must be kept. Files containing intermediary stages of analysis can be deleted unless reproducing the analysis requires an inordinate amount of time or effort.
  4. All files from one individual should be saved in a single folder. Ideally the folder should bear that individuals name.
    1. For short-term laboratory members all experiments can be saved within this folder.
    2. Long-term laboratory members should split their computer files into subfolders, either by year, by project, or by notebook number.
    3. The results from a single experiment should be logged into a folder named with the following format: year-month-day (i.e. an experiment conducted January 14th, 2011 would be saved in a folder titled 2011-01-14). This ensures folders will easily linked to the corresponding entry in the laboratory notebook.
    4. Experiments spanning multiple days should be saved in a folder dated to the first day of the experiment.
    5. For multiple experiments done in a single day, place each experiment in its own folder using a letter suffix to indicate the separate experiments (i.e. 2011-01-14a, 2011-01-14b). Make sure the suffix indicating a specific experiment is clearly logged in the results section of the laboratory notebook.
    6. If there is doubt about whether a file should be saved or not, defer on the side of caution and keep the file.
  5. In some cases files may contain confidential information about patients, donors or experimental subjects. Storage of these files is regulated by the university and by the various funding and governing agencies overseeing the research. It is incumbent on the researcher to ensure this data is stored in accordance with the relevant guidelines. At a minimum, identifying data should be blinded or stored elsewhere. If this is not possible, files must be encrypted using a sufficiently robust encryption algorithm.

Note: this is part of the laboratory notebook and good practices guide.