E-lab notebook concepts and how to use

Transcript

1. E-lab notebook prototype Concept and tutorials ! 4/19/2014 Hiro Kai http://labnote.meteor.com
2. Principles and goals • No missing data • Manage experimental record as logically linked data. • Rigorous and easy management of samples and experiments • Define a experimental protocol prior to experiment. • Samples are automatically categorized by protocols that are used to make them. • Emphasize the realtime-ness of changes in records. • All operations are recorded in time series. • Automatic snapshot.
3. Experiments Samples Basic concept: Experiments and samples are mutually associated • All samples have their origin experiment (= experiment that produced the sample) • A sample may also have experiments that use that sample.
4. Experiments Samples Experiments and samples are mutually associated • All samples have their origin experiment (= experiment that produced the sample) • A sample may also have experiments that use that sample. • Addition of samples does not break consistency of previous data.
5. Basic concept: Experiment “transforms” one sample to another Example: Supported membrane experiment Lipids SUVs Sample Glass coverslip Experiments and samples are considered as an “acyclic directed graph”. Nodes (rectangles in the schematic above) = samples and data Edges (arrows) = experiments Images, FCS data, etc. DLS data, etc. Incubation on glass Making SUVs
6. One experiment consists of multiple steps Example: Making SUVs DOPC Mixing Texas Red Lipid mixture in CHCl3 Evaporation Hydration Sonication Lipid mixture, dried Lipid mixture in water SUVs You possibly want to record details of any of those intermediate samples (rectangles) and steps (arrows).
7. Experiments are hierarchical structure of transformation of samples Any node (samples and intermediates) of a sample graph may be (1) taken out to record details and (2) used for other experiments. Making SUVs Imaging Making SLBs Imaging of proteins on SLB Labeling protein
8. Concept: Multiple experimental runs are carried out with the identical protocol. Protocol as a template of experimental run Making SUVs Actual experimental record (images, experimental parameters, etc.) are associated to these “runs”.
9. (1) Add new experiment and define experimental protocol Workflow of experimental record (3) Record samples and steps Time of experiment Experimental parameters e.g. volume to mix, dilution factor, etc. Sample data (links to data online, etc.) Currently, all raw data (images, spectral data, etc.) has to be saved externally (local HDD, Dropbox, Google Drive, etc.). Every experiment has a protocol that all runs in the experiment will follow. (2) Add experimental runs
10. Screen of e-lab notebook Tabs List of items
11. First thing: define an experimental protocol Protocol panel
12. Defining a protocol This protocol will be a template for recording experimental samples and steps. I’ll show you a demo of how you can make a flowchart you want.
13. Adding runs, and record steps This protocol will be a template for recording experimental samples and steps. I’ll show you a demo of how you can make a flowchart you want.
14. Creating/assigning samples There are two options to add samples to an experimental run: (1) Making a new sample For output samples. (2) Assigning a sample that is made in another exp. For input samples. This will connect nodes (=samples) in the sample graph. New sample is assignable Previous sample is assignable Used for future experiment Input samples Yes Yes No Output samples Yes No Should be Intermediate samples Yes No Possibly
15. Sample types Sample type is a concept useful for keeping consistency of data. Type tree is editable by drag and drop. “Type A is compatible with B” means either (1) A is equal to B (2) A is one of descendants (subtypes) of B. Every sample has a type. Every protocol node has a type. Sample type has to be compatible with the type of a protocol. • This avoids assigning a sample of wrong type to an experimental run. • This makes it easy to find samples from a large set of data.
16. Sample history view Every sample is associated with other samples by the relationship defined by experimental steps. You can easily keep track of all history of the sample you are interested in! • Ancestors and descendants • Experiments that are involved with them.
17. Summary and future directions Once you become familiar, this logically organized sample management will give more efficient and consistent management of experimental samples and protocols. Snapshot and complete logging are desirable features to guarantee reliability of the e-lab notebook. Reporting by PDF, attachment of data, cooperation with Google Drive and Dropbox for raw data storage, etc.