Sixty-Four Free Chemistry Databases Part 21: Exploring Chemical Structures and Data with ChemSpider
Databases of chemical structures become much more valuable when data and structures are linked together. Today's stop on our continuing survey of free chemistry databases and Web services takes us to ChemSpider, a service that like few others embraces the idea of linking chemical structures with data.
From the home page:
ChemSpider is a free access service providing a structure centric community for chemists. Providing access to millions of chemical structures and integration to a multitude of other online services ChemSpider is the richest single source of structure-based chemistry information.
ChemSpider is organized around the idea of a unique chemical structure and various forms of data associated with it. One of the distinguishing features of ChemSpider is the number and variety of forms these data can take.
The entry for diazepam illustrates the rich variety of data ChemSpider offers. Starting from the top, the chemical structure, several properties calculated form it, and several different kinds of identifiers are all displayed. Further down is a prose description taken from the Wikipedia entry. Expanding the box titled "Associated Data Sources and Commercial Suppliers" gives cross-links to other databases and services, including commercial suppliers, safety data, nmr spectra, and many others. Still further down we see links to "Supplemental Information" including GC and HPLC hosted by instrument vendors. Skipping over several other sections, at the bottom we find a dedicated section on "Spectra", from which an interactive IR spectrum can viewed.
Although there's no obvious way to find the total number of structures appearing in ChemSpider, various reports place it in the 20 million+ range.
Beyond its size and scope, another noteworthy feature of ChemSpider is its community curation. Data can be and are curated by individual chemists in their spare time. After signing up for an account, you'll be assigned a role. Unfortunately, the lack of any visible clue as to how to add, remove, or update the data contained in ChemSpider deterred review of these capabilities, and the apparent documentation failed to offer discernable guidance.
ChemSpider has collected and linked a remarkable amount of chemical information, but unfortunately its size, scope, and cluttered interface can make it a difficult resource to understand and use on a daily basis. For example, take the question of how to find all ChemSpider records containing the 2-aminonaphthalene moiety. Starting from the home page, the process consists of eight steps:
- Select the Search option.
- Click on the Structure Search link from the left-hand menu.
- Click on the chemical structure image with a pencil on it.
- From the popup window, click the "Edit" radio button.
- Either click "Run" or "Cancel" on the unsettling security dialog that reads: "The application's digital signature has an error..."
- Draw your structure, one atom at a time, using the editor applet.
- Click the "Accept" button to close the popup.
- Click the "Search" button to begin the search.
I initially forgot to check the "Substructure" button on the search form, which resulted in executing an exact structure query that returned the expected hit. But using my browser's back button (Firefox 3/Ubuntu Linux) put the service into a state from which further search was apparently impossible, repeatedly returning no hits for exact- and substructure search.
Unfortunately, this is not the only instance of a confusing, overly complex, or simply nonfunctional interface in ChemSpider.
ChemSpider is the first service to collect large amounts of chemical structure information, combine it with a diverse array of associated data, and make the result available free of charge to the public. Despite usability issues, it's well worth the time to explore what's possible with ChemSpider.
Kudos
- Massive collection of chemical structures and associated data.
- Community-based curation.
- Integration with other Web resources.
Ideas for Improvement
- Improve the visibility of ChemSpider's capabilities by streamlining the home page.
- Improve the usability of search with a single, popup-free interface for all searches accessible with no more than one mouse click from anywhere in the system.
- Ensure that common actions such as use of the back button work as expected on all major browsers.
Thanks for the feedback, compliments and suggestions for improvement. All good stuff. When I read your review of the system it made me rather proud of what we managed to create over the past couple of years with no funding and with "after hours" time only. The fact that this was all running on only three servers (one purchased and two home-built) has always been rather amusing for me. We received job applications from a lot of people over two years as they thought we were this large organization with a massive architecture and pool of resources in the background.
The community-based curation has been the basis of cleaning up the data and while not complete (for sure!) we've made great strides in cleaning up the data. ChemSpider certainly owes a lot to the community who have helped clean up the data and enhance it with their own data - whether it's more structures, reactions, spectra or other forms of data. We are very grateful.
As you know ChemSpider has been acquired by the Royal Society of Chemistry and as of June 1st we have been working to move the ChemSpider platform off of our servers and onto their servers sitting in Cambridge. It has been a major undertaking but the results are that the system is already more stable, has faster response times and has lots of failsafe systems in place to help support increased uptime. (Previously ChemSpider would go offline when there were cable outages and bad weather systems!)
With the resources of the RSC behind us we are looking forward to knuckling down now "fulltime" and improving the GUI, overall usability and performance of the system. And NOW we are resourced to do so. We will unveil a "slightly" new look and feel at the IUPAC Congress in Glasgow and ACS in Washington but the major changes to interface and usability will come after these conferences later in the year. As always we welcome your constructive feedback!
softwrae please.