To summarise: Chematica is software associated with an extensive database of chemical information. The software is designed with an algorithm which can propose synthetic pathways to produce particular chemicals. Such pathways would, of course, need to be tested through actual experiment, but Chematica's proposed synthetic routes are at least plausible avenues for investigation. A recent upgrade to Chematica allows the software to avoid pathways which are protected by patents.
This means that if one wanted to synthesise a particular drug molecule, but the most well-known pathway to its synthesis were covered by a patent, you could use Chematica to find a synthetic route which wouldn't infringe the patent. On the face of it, the most obvious lesson here is that if you can claim your final product, do so. In some circumstances that won't be possible – if your patent is to a hitherto-unknown way of making a previously-known molecule, for instance – but if your drug is new, putting in a claim to the drug itself in your application is a wise move.
Of course, the patent in question might target that particular synthetic route for a good reason. Perhaps the end compound is already known from previous prior art – but the particular synthetic route patented has particular advantages. It might be more efficient, or require cheaper starting materials, or be safer in terms of reaction conditions and the by-products produced, or be more convenient to carry out in a pharmacy or home setting – there are any number of reasons why a particular reaction pathway may offer an advantage over others – an advantage which offers a patentable contribution to the field. Chematica might help you find a different route to the target drug – but there is no guarantee that this route would be a commercially viable one. If the protected route is much more efficient and/or cost-effective than the alternatives, the patentee still has an advantage in the market.
Then again, through the use of Chematica you could stumble across a synthetic route which has its own distinct advantages – enough so that, in principle, it might be patentable by itself... but wait a second. If the use of Chematica and similar software becomes a standard part of the chemist's trade, it would arguably be obvious to the skilledlausible person seeking to synthesise a particular drug to use Chematica to research potential synthetic pathways. In fact, the more powerful the Chematica software becomes, the more reason there is for the skilled person to use it. Supposing Chematica included output on, say, the expected efficiency of the synthetic routes it proposed, or the expected costs, or other such data?
The more such information Chematica provides, the hard it is to argue that the synthetic routes it proposes don't come under the umbrella of "obvious to try" – and therefore unpatentable. It's a curious side-effect of patent law's focus on the notional "skilled person" that as new tools become available to the skilled person, the more they become capable of – and since to get your patent you need to demonstrate more inventiveness than the skilled person, the more capable the skilled person is, the more difficult it becomes to get a patent.
Therefore, it's worth noting that the net effect of tools such as Chematica may be quite negative for patents and patent applications whose claims hinge on a specific step in chemical synthesis – not only does Chematica make it easier to circumvent that step and evade your patent protection, but if the step is one which Chematica would have proposed, it might not be considered all that inventive in future.
Conversely, a patent to a particular drug molecule itself, defined without reference to the synthetic route (and perhaps instead by reference to, say, the medical use of the drug molecule) is safer from this type of circumvention. Once everyone knows the intended destination, it's possible to plan all sorts of different routes there – but it's another matter to discover that there is something worthwhile at the destination in the first place. Similarly, if one were to discover a radically new chemical process or reagent – one which made possible a class of synthesis which would not otherwise be possible – or wouldn't be commercially viable – via existing tried-and-tested routes, then that could offer excellent protection; Chematica could, in fact, be the patentee's friend if its use reveals that there's a particular synthetic "bottleneck" to producing a target molecule, and your patent happens to cover that "bottleneck" as well as the target molecule.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.