I recently wrote about our software business model and how mcule.com can serve as an online, integrated drug discovery platform for different users by offering attractive, long- and short-term software tool subscriptions. In order to get the most out of these tools, they need to be integrated with a high quality molecule database. It is equally important to make sure we have a properly prepared molecule database optimized for the searching/screening tools.
Chemical correctness. We have spent many months on our molecule registration system, to filter out problematic structures retaining the highest quality compounds only. You can find more information about our registration system here.
Purity. 90% but rather 95% is the typical required minimum purity for screening compounds. For some natural products, this limit might be a little lower (80-85%). These values are pretty standard and most chemical suppliers fulfil these criteria. Stereochemical purity is, however, a different story. Suppliers typically send their available compounds in SDF format, but unfortunately stereochemical information is not properly stored in these datafiles. Furthermore, the SDF v2000 format is unable to handle the stereochemistry of several cases. We therefore ask specific stereochemical questions from our suppliers, to make sure their molecules are processed properly.
Identity. NMR and LC-MS spectra should be attached to the shipment or at least accessible upon request.
Acquisition rate. This heavily depends on whether the screening database is up-to-date, how quickly the screening can be completed and whether we have instant availability on the individual compounds when ordering. Worst case scenario: company has a chemical supplier database updated annually, screening takes few weeks (e.g. to screen 1 million compounds by structure-based docking), and they send the IDs of virtual hits to chemical suppliers. Acquisition rate might be <50%. On the one hand, this is a waste of resources (human and computational) spent on the evaluation of non-purchasable compounds. On the other hand, these resources could have been used for compounds that were not included into the screening database because it was out-of-date. A frequently updated database of purchasable compounds integrated with the screening tools themselves is therefore a much better choice. Having instant availability on the compounds when ordering could further improved the acquisition rate close to 100%.
Delivery time. These molecules are intended to serve as interesting chemical starting points for new/early phase drug discovery projects. Therefore it is crucial to get these molecules delivered quickly. New and early stage projects try to exploit all available hit sources. To be able to compete with e.g. in house high-throughput screening, and to have an impact on project decisions, hits coming from external resources should ideally arrive before the in-house experiments starts. Otherwise thoughts of medicinal chemists will be primary tailored by the hits already sitting on their table. Everything else will be sitting on bench. First hits will have priority, and therefore quick delivery plays a crucial role in this competition. Typical industry standard for delivery is 2 weeks.
Single package delivery. Molecule procurement can be very painful. It is definitely not something a scientist would like to deal with. It is therefore very important to offer the possibility to deliver compounds produced by multiple chemical suppliers as a single package to the door of the customer. Instead of negotiating with the representatives of chemical suppliers and deal with custom clearance a scientist should do what he/she is best in - science.
Price. Another important aspect of course is the price of the compounds. Since several compounds are typically offered by multiple suppliers, it is important to find the best deal. Besides the price per compound, other factors determining the final price include the number of compounds ordered per supplier, minimum order fees of suppliers, delivery costs, delivery format and required delivery time. It is important to have a good algorithm that compares all possibilities and finds the best choice for the customer.