Trial Synthetic Procedure for Chemspider to Find...

Published by MatTodd on 12 February 2009 - 9:56am


general open research

There has been an interesting conversation over at Friendfeed about the value of centralised vs. decentralised synthetic procedures for chemistry. i.e. should we try to create a store of chemical reactions, or ought we to be expecting decent search tools to be able to find these for us, wherever they are. From my perspective, both have value. Naturally what we need are tools that understand descriptions of reactions that are semantic (plain English), and which can understand various forms of data that go along with compounds.

So I thought we'd give it a try with a random sample procedure, the conversion of tetrahydroisoquinoline to its Boc-protected form. Below is a simple procedure for this, written as if we were submitting this to some web archive of procedures. Also attached are some data for the product - .dx files for JSpecView-ing the proton NMR spectrum, plus a pdf of the same. There are gifs of the reaction itself, plus one of a TLC of the reaction. I'll shortly upload some other things below. I have no idea what any search tool will make of such a mish-mash.

Antony Williams can hopefully unleash ChemSpider, and extract information from this? I also want to try to upload the same data directly to the relevant page on Chemspider, to see if I can just deposit it there. Which of these two approaches is preferable?