Tuesday, June 26, 2018

The Disconnection Approach - Automated!

In my time as a synthetic organic chemist the most important advance in the field was definitely the introduction of searchable databases such as SciFinder and Reaxys. Life before these involved spending days on end in the library flipping though dusty tomes of chemical abstracts and Beilstein. And at the end you weren't even sure if you missed something of critical importance. The introduction of open access LC-MS and high field NMR has also had a big impact for me by speeding things up considerably. However, besides these milestones I think that I have pretty much been doing chemistry the same way >25 years. Anyway, what I am getting at is "what will the next BIG thing be?" It's been a considerable time since we had a major technical breakthrough for the synthetic organic chemist. My colleagues and I have been discussing this for more than a decade and now I think I am spotting the next big thing - Chematica! Chematica was developed by Polish scientist Grzybowski and has been around for quite a while. Basically it is a computer programme that disconnects your molecule and suggests ways for you to synthesise it. What's changed since I first heard mention of Chematica ~5 years ago is that apparently now it works the way you would like it to. Recently Grzybowski and co-workers published an impressive paper where they synthesise 8 very different and rather challenging molecules. The abstract from the paper nicely summarises the achievement:

"Multistep synthetic routes to eight structurally diverse and medicinally relevant targets were planned autonomously by the Chematica computer program, which combines expert chemical knowledge with network-search and artificial intelligence algorithms. All of the proposed syntheses were successfully executed in the laboratory and offer substantial yield improvements and cost savings over previous approaches or provide the first documented route to a given target. These results provide the long-awaited validation of a computer program in practically relevant synthetic design."

You really should read this paper. These are not simple syntheses and would have taken quite some time to come up with if at all. Now it is still early days and Chematica is only for those with deep pockets. I am personally waiting for a quote right now and am very curious to see exactly how deep my pockets need to be. BUT it has started and I believe that now it's just a question of time (I'm guessing  less than 10 years) before you simply hit the disconnect button in ChemDraw, look through the suggestions that appear on your screen and pick the one that you like best.
And there is more. Another game changer is already here and everyone with a PC can do this: Machine Learning. My next post will be on this topic that we are already using to great effect at my work place. D!



No comments: