Ali Baba is pretty neat.
Ali Baba parses PubMed abstracts for biological objects and their relations as discussed in the texts. Ali Baba visualizes the resulting network in graphical form, thus presenting a quick overview over all information contained in the abstracts.
Perhaps the best way to explain what it does is with an example:
A patient with cough is treated with codeine. He becomes unresponsive after a while — what is going on?
The query entered in Ali Baba was “codeine intoxication”.
Ali Baba shows the relationship between codeine (marked in the graph with blue frame), cough, morphine, and poisioning. Poisioning is also connected to morphine and CYP2D6. The solution thus is that codeine is bioactivated by CYP2D6 into morphine, certain patients show an ultrarapid form of this metabolism, which leads to a life-threatening intoxication (see Gasche et al. (2004)). The connection codeine->CYP2D6->morphine is directly visible in Ali Baba.
To make this example a bit more concise and focused on the central problem, we used the minimum degree filter (in the menu, see Preferences|Filter preferences). This allows to remove all nodes without any neighbors (unconnected nodes) using the ‘Degree’ slider on the bottom panel.
I’m not sure Ali Baba will be of any use to me, but it really is neat. Please note: Ali Baba requires Java 1.5 or higher.
Other posts about third-party PubMed tools:
- FABLE (3rd Party PubMed Tool)
- Managing Medical Literature on a Mac: iPapers, Papers, Sente, BibDesk
- Notes on ReleMed
- PubMed Gold
- PubMed Reader
- For MedLibs who use Macs: iPapers
- PubMed2Connotea / PubMed2CiteULike
- More notes on BioWizard (Digg for Medical Literature, Part 3.5)
- More Alternate PubMed interfaces via Journalology
- BioWizard Enhancements: ‘Digg for Medical Literature’ Part III (Edited)
- Some Alternative Interfaces and Mashups for MedLibs
- BioWizard: The start of ‘Digg for Medical Literature’?
- Article on HubMed