CElegans MySQL Database

[InterIntelligence Research]

As promised in the Data meeting, the semantic data base is now in a MySQL form:

 

 

Database Name:	mysql_31129_celegans
Database Server:	my01.winhost.com
Database User:	openworm
Database Password:	openworm

 

I have not changed on the application to look at this data base yet but will over the next couple of days.

 

Thanks

 

Tim Busbice

[Dan Knudsen]

Thanks, Tim!

I look forward to getting my connection set up and taking a look.

-Dan Knudsen

[Stephen Larson]

Hi all,

   I just created a super simple example of reading this via Python over here:

http://nbviewer.ipython.org/6693044

   I'm thinking it would be pretty straightforward to be able to start querying this via semantic graph queries if some Python code were written to munge this into an RDF graph.  Here's some ideas on how that could be done:

https://rdflib.readthedocs.org/en/latest/intro_to_creating_rdf.html#adding-triples

Thanks,

  Stephen

--
Visit us online at:
http://openworm.org
http://blog.openworm.org
http://github.com/openworm
http://twitter.com/openworm
https://plus.google.com/s/openworm
---
You received this message because you are subscribed to the Google Groups "OpenWorm-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openworm-discu...@googlegroups.com.
To post to this group, send email to openworm...@googlegroups.com.
Visit this group at http://groups.google.com/group/openworm-discuss.

To view this discussion on the web visit https://groups.google.com/d/msgid/openworm-discuss/be01e2e9-769f-486a-a8fd-71f886316953%40googlegroups.com?hl=en.

For more options, visit https://groups.google.com/groups/opt_out.

[InterIntelligence Research]

Very cool - I know what I'll be doing tomorrow :-)

 

Tim

To view this discussion on the web visit https://groups.google.com/d/msgid/openworm-discuss/CAHxRZu6%2B9-65jw4C7Xq%2BYw6QRbk2rMyTfuUZnZ7DpJPHcnBf%3Dg%40mail.gmail.com.

[Stephen Larson]

Hi all,

There's some new activity on the issue tracking this over here: 

https://github.com/openworm/OpenWorm/issues/143

We have the beginnings of a script that opens the OpenWorm semantic database, generates RDF, and does a graph query (SPARQL) to pull information out a single neuron.  There's an iPython notebook that successively walks through examples posted online.

If folks like this, we may be coming close to finally finding the ideal integrated format for the information we have long held in spreadsheets, then google spreadsheets, and then this database.  The final example script in the notebook above could now rapidly be adapted to be the input to RegenerateConnectome.py (issue 140).

Thoughts are welcome.

Best,

  Stephen

To view this discussion on the web visit https://groups.google.com/d/msgid/openworm-discuss/CAJnfD%3DbgTYYf8ApUD1sTNkg4DxZS4AqN3DBqzG-x6fq5TcDGCQ%40mail.gmail.com.

[Padraig Gleeson]

Hi,

Great work Stephen! This will indeed be very useful for feeding data to and validating the NeuroML model.

Sorry to keep banging on about this though... Almost as important as what the data is that's going in to the model is where it's come from... It's very tempting for us as hackers to get structured data and put it into our models, but if you can't link those numbers back to the original experimental data it will be hard to convince people further down the line that it's a physiologically detailed model. For any query that returns info that there are N connections between cell A and cell B, I should be able to do queries like "Who added this to the database?" "Which lab produced the data?" "Where is a description of the experimental preparation?" "Has data been used/described in a publication?".

As a first step it may require adding a top level entity to the database "Data Source" with a free text field (and URL to ext resource) and maybe a field "Data submitter" (then all relationships can point to one or both), but the longer we take to address these questions, the more difficult it will be to add these features in future.

Padraig

To view this discussion on the web visit https://groups.google.com/d/msgid/openworm-discuss/CAHxRZu6PA13o1gnZcLtT_bruz3y6W%3Dpydz2z9tP3XxWrNR%2BvjQ%40mail.gmail.com.

For more options, visit https://groups.google.com/groups/opt_out.

-- 

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.gl...@ucl.ac.uk
----------------------------------------------------- 

[Jim Hokanson]

Agreed!!!

It also seems like it would be worth breaking up data source a bit further to into the other things that Padraig mentioned (instead of just leaving a text description).

Perhaps:

Data Source Type:

- publication

   - identifiers????

- derived calculation or inference

   - other data source type entries that were involved

   - a description of the derivation

   - gist link?

- experimental data

   - source <- ??? not sure how to define this in a "nice" way - public DB, lab, ?

   - published? <- note, the data values might not be explicitly published but the data set itself might be, it would be good to decide which goes 

   - publication source <- reference to above data type if published is true

These could also be an array:

neuron x resting membrane potential:

1) -80 mV - publication 1

               - publication 4

2) -85 mV - publication 2

3) -83 mV - publication 3

This of course gets more complicated if instead you have:

neuron x resting membrane potential at temp 2 (where temp 2 is not "normal")

Anyways, I'm getting a bit carried away but I very much agree with Padraig's sentiments.

Jim

[InterIntelligence Research]

Can we get a data meeting setup and discuss what and how RDF can help us, how we can use going forward, what if anything we need to do differently than what we are doing now?

 

This is very interesting - I just want to make sure I (we) understand what it is and make some decisions on how we can move forward.

 

Thanks

 

Tim

To view this discussion on the web visit https://groups.google.com/d/msgid/openworm-discuss/CAHxRZu6PA13o1gnZcLtT_bruz3y6W%3Dpydz2z9tP3XxWrNR%2BvjQ%40mail.gmail.com.

[Mike Vella]

Padraig does make a very good point, data is less valuable without easy access to the relevant metadata. However, Tim and Stephen's contributions here are certainly the first steps towards this.

To view this discussion on the web visit https://groups.google.com/d/msgid/openworm-discuss/524971F4.5090305%40ucl.ac.uk.

[Stephen Larson]

Hi Padraig -- yes this is a very good point.  Couple things on this:

1. Tim started putting provenance information like what you ask into the OpenWorm semantic database already.  There are relationships that have citations already.  We talked a little bit about how to expand this in the last data meeting
2. The mechanism in RDF to do this looks like it should be using named graphs.  Here's an example of how that works.  I'm working up an example that combines the example I started with this model of provenance.

Best,

  Stephen

To view this discussion on the web visit https://groups.google.com/d/msgid/openworm-discuss/CAGKgUfVTir%2BYXLzuzgc__GLEHyMLBFQ-Ee3zYo2dcPnTuAdazQ%40mail.gmail.com.