SimBioSys Blog

This week we held two design strategy meetings with Life Science companies from the East Coast. One was focused on a new de novo design package that we are working on while the other was for ARChem, our retrosynthetic analysis software platform (http://www.simbiosys.ca/archem/index.html). I’ll comment on the ARChem meeting I led while Zsolt can comment separately on the de novo design meeting he led.

When designing our software solutions we engage our users in providing feedback to us regarding needs, their biases in terms of scientific approaches and their thoughts about improving workflow, usability and algorithms etc. ARChem has been used in large pharma for about 3 years and our recent installations in new companies, and the resulting feedback from the users, has us focused on the next release cycle for the product. With this in mind we chose a different approach to gathering input.

We brought together scientists involved in the original design for ARChem (and therefore experienced users) as well as chemists who had recently trialed the system. We were interested to hear in a public forum, with issues regarding proprietary approaches put to one side, what would they like to see implemented in ARChem to satisfy their needs and move ARChem one step closer to being an ideal platform for chemists to perform computer assisted retrosynthetic analysis. By the end of the meeting we had rank ordered over 30 specific requests that came up in the meeting and the collective attendees had agreed to the primary issues to address for their needs. Some of the requests we had not even considered prior to this meeting and it was definitely one of the best uses of time and a great design session based on user needs…one to be repeated. We are off to work on the outcomes of the meeting and will keep you informed here of our progress.

posted by Aniko

Some of you are likely aware of ARChem, our retrosynthetic analysis software (http://www.simbiosys.ca/archem/index.html). ARChem is the result of 4 years of development and results from a collaborative project with a major pharmaceutical company. Since then we have delivered the system to a number of other companies and we have recently submitted a publication to JCIM (http://pubs.acs.org/journals/jcisd8/index.html). It should be in press shortly. We presented on ARChem recently at the ACS meeting and a copy of the presentation is here.
What we’ve been up to recently is delivering on the needs of some of our users in terms of integration to latest ChemDraw ActiveX component (http://www.cambridgesoft.com/software/details/?ds=2&dsv=92), expanding the list of starting material databases supported by ARChem and, our most exciting news, working with the entire Beilstein reaction database (http://en.wikipedia.org/wiki/Beilstein_database). I had reported previously on the fact that we had been working with the Beilstein database (http://www.simbiosys.ca/blog/2008/05/30/29/). Since then we have reached an agreement with Elsevier to utilize the entire reaction database in order to train our clustering algorithms. More about this in the future but an example image is shown using the Beilstein Database below. Notice on the left of the image that an example reaction from the Beilstein database is displayed.

Over the next few weeks you will see us blogging about the future development of ARChem. We are about to have a roundtable meeting with thought leaders from large pharna regarding the future development of ARChem and will be focused on the outcomes of this meeting to guide our development during our next coding cycle.

in the Chemistry World, Oct 2008 issue: