Month: May 2012

Storing chemical data with noSQL (MongoDB): a document-based approach

It’s been awhile since I last had the task of maintaining a chemical data warehouse using an SQL relational database. That’s not exactly a coincidence: I put in my time and did a lot of work with Oracle and MySQL back in the day, but my takeaway conclusion was that the transactional table-based systems are profoundly unsuitable for scientific data. The recently popular wave of “noSQL” database systems (such as MongoDB) are, on the other hand, quite a natural fit.

Some of the recent developments with Open Drug Discovery Teams, and content hosting for web sharing, have necessitated that the molsync.com server trade in its former stateless purity, and run a database.

Continue reading →

Chemical data hosting on molsync.com

As a follow on from a recent post, the idea of sharing chemical data by anonymously uploading it to a server, then making it available in value-added ways using a wrapper service, has been embellished. For the last year or so, the molsync.com server, with its all-original cheminformatics software stack, has provided services for various apps from Molecular Materials Informatics, such as rendering content as Microsoft Office formats, parsing ChemDraw CDX files, matching scaffold-substructures to SAR Tables, calculation of physical properties and tautomers, and enabling sharing of data from the Dropbox public folder using the MolSync app.

Now it goes one step further: it is possible for the apps to upload chemical documents (molecules, reactions or datasheets) directly to the molsync.com server. The uploads are anonymous and public, and are stored in a database on the server. On successful upload, they can be accessed via the returned ID number. For example, check out:

http://molsync.com/share?mol=1

http://molsync.com/share?ds=1

Continue reading →

Sneak preview: 3D molecule viewing in Open Drug Discovery Teams app

The Open Drug Discovery Teams app will soon acquire a new feature: viewing of 3D molecules from within the app. It will be the first app from Molecular Materials Informatics to make use of recently developed OpenGL code for molecule rendering and visualisation in 3D. Click on the thumbnail on the right for a very short demo, which shows ODDT in action, opening up a factoid-tweet that provides as its payload a 3D conformation of aspirin.

Continue reading →

Tweeting chemistry from mobile apps: one less step

A new feature will be making its way to the Mobile Molecular DataSheet, and other apps, soon: tweeting out chemical data (molecules, reactions, datasheets) without having to login to a repository for storing the source data. All that’s required is authorisation of your Twitter account, as configured within iOS.

To see a preview of how it works, click on the image to the right, to watch a short screencast.

Continue reading →

Venturing into the world of 3D

Over the last week or so, I’ve been spinning up some test code and infrastructure for supplementing my mobile apps with 3D viewing capabilities. This involves adventuring into the wild world of OpenGL ES, which is quite a maze of toolkits, meta-toolkits, versions, subsets, deprecated functions and optimisation gotchas. Nonetheless, I now have a widget capable of displaying 3D MOLfiles, and thus far allowing finger-dragging to rotate, as you can see if you click on the image to the right and watch the short screencast.

Continue reading →

UI tweaks: corner tabs and general configuration settings

The editor panels for the MMDS app (structures, reactions and datasheets) have been augmented to include new buttons along the top corners, which can be used to close the editor. These are analogous to buttons in the main bank for close and save and close without saving. This addresses a minor but persistent inconvenience: previously, the main command bank had to be open in order to exit from any of the editor panels. Continue reading →

Sneak preview: SAR Table matrix view

The SAR Table app is currently undergoing surgery to add a major new feature: the matrix view, which allows two columns to be plotted against each other, e.g. R1 vs. R2. Cells at the intersection of two values are plotted with a colour/gradient that is indicative of the response, i.e. the activity. The sample snapshot to the right displays several activity values using the “heat map” colour scheme.

The new feature is in its early stages, and the visual feng shui will certainly mutate before the next release, but the core functionality is operational. Stay tuned for the next major version.