In my previous article on synthetic biology, I briefly touched upon the general findings on the application of synthetic biology to small molecule drug manufacturing, obtained from my Cambridge University Master’s industrial placement project at Cambridge Consultants. In my research, one of the goals was to rationally identify opportunity areas for SynBio in drug manufacturing, a.k.a. develop a method to be used to determine specific drugs or classes of drugs that could highly benefit from a SynBio-based manufacturing process.
Prices of drug: a starting point for understanding the difficulty in manufacture
The identification of opportunity areas was firstly based on the classification of drugs into 44 compound categories, such as alkaloids, steroids, and terpenoids and on the price information for several drugs in those classes. The assumption that I took was that the difficulty in manufacturing drugs is mirrored by the cost of manufacturing which in turn is related to the price of the drug, which is therefore related to the opportunity of applying synthetic biology (it would not make sense to apply SynBio for a class of compound which is very easy to manufacture),. In my assumption, this was valid only when referred to generic drugs, as the prices of innovative on-patent drugs include a premium which is generally not reflective of manufacturing costs.
Therefore, in my analysis, I used prices as proxies for manufacturing cost/manufacturing difficulty only for drugs available as generics. The idea then was to compare the pricing information against a further parameter in order to screen for opportunity areas on the bases of these two parameters.
Assumptions for selecting opportunity areas
In order to predict which class of drug represents the biggest target area three hypotheses were made and then tested:
- Larger molecules are more complex, and hence harder to make and more costly as the bigger they are the more steps are needed to get from the precursors to the final molecule.
- Certain particular drug classes would be hard to make.
- Molecules with more chiral centres would be harder to manufacture as the chiral chemistry is often complex and expensive.
Where to get the data and what to do with it?
I took the information on the price of 150 generic drugs from the American drug price database “ASP Drug Pricing Files” which contains the average selling prices of marketed drugs, used for referencing reimbursement in the United States. With this information, I then built three graphs aimed at identifying interesting classes of compound. In the first case the price information (used as a proxy for manufacturing cost) was put against the size (molecular mass) information, with the objective of isolating outlier drugs, namely compound that were small (so expectedly cheaper) but exceptionally expensive, to then deduce which classes of drugs were more represented within this group of highly-priced/low weight drugs. In a second analysis, the price information was related to the number of chiral centres for each molecule. The objective was to understand if the number of chiral centres could somehow predict the price of the drugs. For the last analysis, only the price of drugs was taken into consideration. The minimum, maximum and average prices were plotted for each class to understand which class is generally more expensive than others.
Therefore, those three parallel analyses were performed to easily and rationally identify “opportunity areas” for SynBio in small molecule drug manufacturing. In the next article of this blog series, I will report the result of the analysis and what arose from it.