Joining Forces: Drugs in the Cloud

Tell us about your project…

Eric Fotheringham: For some years, we have had a collaboration with the University of Copenhagen (Petur’s lab) focused on mass spectrometry (MS) applications in forensic toxicology. More recently, this has culminated in us asking him to participate in our Centers of Innovation (COI) Program, working on the monitoring of novel psychoactive substances (NPS).

Petur Dalsgaard: The field of NPS is very fast-paced; there are one or two new compounds hitting the streets every week – so getting the information into your target library is one of the biggest challenges. There is a risk that by the time you finally have the reference standard in your system, the compound has fallen out of favor. So we’re trying to build new databases to get ahead of the game. We now have a cloud database called HighResNPS (www.HighResNPS.com); every time a compound is detected somewhere, we can get the reference values to use in our QTOF-MS system. If members find a new tablet or designer drug in Finland, for example, then they post it in the library, and I can use their values in my system, regardless of vendor – without any reference standard. Each member has different instrumentation, but they can still use the information in the database. At the moment, it’s mainly the European countries who are involved, but we’re trying to get as many labs as possible to join in. It’s free, but everybody has to contribute findings. You could describe it as crowd-building – if everybody contributes, then each party gets back what they need many times over.

We are hoping that building a library this way will be faster than the European Monitoring Centre for Drugs and Drug Addiction (EMCDDA). Databases such as mzCloud (www.mzcloud.org) are really big now – there are 500 NPS compounds (out of a total of 6,500 compounds). The challenge, of course, is how to use these cloud databases...

How did you build the partnership?

PD: I guess it really started ten years ago on our side, when the University of Copenhagen purchased a Waters UPLC TOF-MS system; I was hired to get it up and running – and to conduct forensic toxicology screening. One of the key people I worked with at Waters was Michelle Wood – she did the instrument training and I continued to correspond with her afterwards. Later, I got to know some of the programmers, and it evolved from there. We now have a very close relationship with the company.

EF: The fundamental premise of the COI Program is to carefully select and partner with thought-leading scientists around the world who are already using our technology, whether that’s a simple sample prep product, a piece of software or a full-blown LC-MS system. We reached out to Petur for a collaboration because he has access to an incredible range of samples – from police departments and so forth around Europe; there are few people who could claim the same! Petur has now run many of those samples on our MS-based systems, which has given us some very interesting insights. Not only has it enabled us to further develop the instrumentation but also – even more interesting to us – the software that collects, processes and archives the datasets. It’s also enabled us to expand our UNIFI Scientific Information System to include libraries of related drug compounds and their metabolites.

What are the benefits of collaboration between academia and industry?

EF: That’s a question I’m often asked: “What does the scientist get out of the relationship?” The key word is again ‘access’. In April of 2016, we gathered our Program participants together in the UK for a set of workshops. We had them interact with our senior management team – including our CEO and the Senior VPs – in an open forum. We discovered that our partners greatly valued that access – both to the people in our company, and to new technology.

PD: It’s true – the benefit of collaborating with Waters is that you have direct contact with the people who are programming the software or making key decisions. I have worked with Waters for a long time on software development for MS instruments; in fact, we have been working with them on UNIFI since 2011, doing beta-testing on each new version, implementing it and feeding back to them. I subsequently have a very good relationship with the programmers at Waters. We might feedback some aspects that we would like to see in the next version of the software, and that can be a huge benefit – it’s almost like getting custom-made versions for our lab!

Waters have also lent us a new instrument – an ion-mobility QTOF – and we’ll be putting it through its paces to see how it performs compared with the older-generation QTOFs.

Although manufacturers have great ideas about how to build the instruments and the software, the truth is that, unless they try it out in the field and see how it works, they won’t know how useful a design feature is, or if they’ve omitted something that should be there. We universities have the ability to do more in-depth work than instrument makers...

EF: I completely agree. We were interested in the science – there’s a world of samples that we can’t get our hands on directly, and that was a unique benefit for us. The benefit to Petur is being able to give his input, so we can provide him with the software that he really needs for his lab.

What challenges do you have to tackle?

PD: One of the challenges for Waters is the time it takes to develop everything – to get everything programmed, and wait for the next version. If I ask for a software change today it could be six months to a year before we see it implemented; it’s not something that can be written in a day and that you can download and run the next day. One of the challenges with the database is the maintenance – it’s ‘homemade’ at the moment, so we need funding to improve it and make it more professional. It would also be great to centralize support for all the labs rather than having someone from each group updating it.

EF: We always have to be mindful of the workload of our partners. When you visit Petur’s laboratories and see the throughput – the sheer number of samples – it’s an eye-opening number! At certain times, they are operating at full capacity, and that’s not the time to insert a new request to run a new version of software. It’s easy to overreach or over-ask, but I think that can be mitigated by good planning upfront and setting clear expectations.

What makes a collaboration effective?

EF: Trust. As I mentioned, when we choose a partner or enter into a collaboration, there’s already a relationship there. It may start with a transaction – they buy something from us – but that’s very basic. We want to develop that into a much closer partnership. I think when there is regular contact, and the benefits are clear to the partner and to us, it’s got a great chance of working out.

Want to see your collaboration featured in these pages? Get in touch at [email protected]