As some of those who are on LinkedIn, and specifically the members of the discussion group ” Drug repurposing – reprofiling – repositioning,” are already aware we are currently debating how a database on drug repurposing candidates could best be established. To be exact, this database would compile data on active ingredients of discontinued drugs or drug candidates. This is only a subset of all potential repurposing candidates because it is almost always possible to identify new potential uses for almost any pharmaceutical compound (its only a matter of searching hard enough). But this database will only include compounds that have failed in their initial development, or have been discontinued for other reasons. We envisage it as being very comprehensive, combining development information, medicinal chemistry, pharmacology, and patent information. And it will be exclusively based on public sources.
Now two questions could be raised about that. First, aren’t there enough pharma databases already from which this type of information could be extracted? The answer to that is that no online database exists with this particular focus on repurposing in terms of compound selection and subsequent data processing. Although it might be possible to extract this type of information from all-purpose pharma databases, the particular set of data that interests the potential drug repurposer would not be there, or not formatted in the correct way.
Second, how could this work with only publicly available data? The answer to that is Open Source Intelligence (OSINT), which H.M. Pharma Consultancy has been using for years. (See the August 2010 blog entry here.) An amazing amount of information can be directly deduced, and much more can be indirectly inferred with a quantifiable degree of likeliness, by combining and assessing open sources in the right way and by doing what is known as traffic analysis. (See for instance, the Recorded Future website. OK, this is for politics and economics but you will get the gist of it.) Information collation can be federated or even crowdsourced; information processing can be in part automated. The tools exist, also for the chemistry.
What remains now (funding left aside) is how we are going to structure it; and this depends to a good part on how we decide the Failed Drug Database would be accessed. Getting an institutional non-profit partner and makeing the database Open Access would be one way; the other is to get investment that can be monetarized by making at least the deeper strata of knowledge in the database subscription-only.
Some way or other, this is going to go forward. Stay tuned, even if you don’t here about this for some time.