[Rutherford]

Does anyone know a good source for practicing retrosynthesis?

Something like this, where solutions and guidelines are given: http://www.chm.uri.edu/mlevine/chm427-2013/retrosynthesis%20solutions.pdf but with a wider range of applied reactions (and correct stereochemistry of trigonal carbon atoms  )

[Rutherford]

I'm pretty desperate so could anyone post a random molecule of moderate complexity for which I could try to write a retrosynthesis scheme?

[kriggy]

I´ll try to look for something, I have some interesting molecules Im working on but since they are not synthesized yet, I would rather not share it this way.
Try this:

[Rutherford]

Here is my attempt for the right side molecule (the left side one would have a similar synthesis, just with other protection groups):

The formation of the Grignard could require the protection of the OH groups. I ask for feedback.

[orgopete]

When I was an undergraduate, we were frequently asked to synthesize molecules from 'air, fire, and water'. Little did I know that as an industrial chemist, I had to know this also. Basically, the cost of all chemicals depends on it.

If you teach organic chemistry, you can quickly discover that the only reactions you can predict the starting materials for are the ones you can draw the products of. My education pre-dates the popularity of the term 'retrosynthesis'. We still did them, but just not with that term. So, I am not in love with retrosynthesis for the sake of retrosynthesis. I still fall back on you would find it difficult to predict the starting materials for reactions you don't know. In my opinion, you should use retrosynthesis to overcome boredom or in an attempt to make a very simple problem harder, such as Diels-Alder reactions.

You can learn one kind of retrosynthesis from the Aldrich catalog. You may find another kind in which a professor searches for a molecule that can fit a reaction he or she has developed. At this point, we come to what should one do when then are given a new class or field of chemistry. What I always found useful was to learn which reactions work the best for this class. I would read reviews.

Short answer to question, check out a review, predict how to synthesize the target, and then read how it's done (or just read how it's done and then imagine what can be made with that chemistry).

[Rutherford]

In my opinion, you should use retrosynthesis to overcome boredom

Yep, I find doing retrosynthesis very amusing. The hard part is to choose which molecules to synthesize, as when I search, I already have some thoughts about possible reactions and I easily discard molecules to accept the ones that suit me most, that's why I need some tests or problems with random molecules.

PS The intermediate for Pictet Spengler is drawn incorrectly.