Since I returned from Hong Kong I have been researching a prospect that would draw upon experience, expertise and technology from Banking and FinTech, and apply to it solving some of the core challenges in the Biotechnology sector. “Critical Issues in Bioinformatics and Computing” published in the journal Perspectives in Health Management as long ago as October 2004 outlined the immediate and future problems that Biotechnology faced at the time and unfortunately still faces today.
While one can’t deny the huge amount of progress in Biotechnology and Bioinformatics, especially in the last decade, and specifically in the field of gene sequencing with Next Generation DNA Sequencing, handheld sequencers and personalised health, to a large extent these advances have only exasperated the issues surrounding data and information privacy, sharing and analytics. The sheer volume of data; global, regional and national regulatory regimes; lack of standardisation, legacy technology and the pace of growth are just some of the factors that conspire to reduce the efficacy and efficiency of cooperation and collaboration in the field.
To anyone working within Banking and FinTech these problems may sound very familiar and the good news for Bioinformatics is that the people, processes and technology required to make a serious attempt at resolving these issues exist today. Arguably all that is required is to bring together the right people from academia, public health and the private sector, under a coherent strategy to break down and tackle each issue utilising state of the art technology. I don’t wish to underplay the size or difficulty of the challenge but from my perspective, many of the problems that currently prevent major opportunities from being seized and realised in Biomedicine have proven solutions with their origins in Banking, Financial Services and Payments sectors.
As a technologist, reviewing the software employed in Bioinformatics today, what I believe to be missing is a community-driven platform that can incentivise the sharing of data, and the algorithms used to mine that data. Of course, there are the large Bioinformatics Databases (NCBI, EMBL-EBI, DDBJ) that share data between each other and which offer a wealth of data and analytics to researchers, however, creating fully automated analytic circuits that pipeline data between algorithms can still require tremendous effort using existing tools.
Ultimately the task falls to individual research teams to create the algorithms and frameworks they need and their output can be difficult for other researchers to access and use. Researchers spend more time on setting up and configuring their IT platforms than they do in interpreting the results they provide. Add to this a large dose of square wheel re-invention and opportunities for a better approach become obvious. The likes of NCBI, EMBL-EBI and DDBJ are amazing resources for researchers and private industry alike and while they facilitate innovation they are not its source in Bioinformatics.
Creating a platform that incentivises and simplifies the exchange of information is, I believe, the key to generating and capturing sustained value in Bioinformatics. Just such a platform is what I hope the Biomedical Algorithmic Exchange (BIOMEX) will deliver.
Selecting the right technology stack for such an exchange will be a fundamental factor in achieving early success. In order to meet the demands of managing a wide range of use cases and to streamline the onboarding (listing) of data and algorithmic assets on the exchange, containerisation and a hybrid cloud architecture are mandatory requirements as is the ability to identify and select the best execution venue (multi-cloud spot price IaaS consumption).
Monitoring and metering access to exchange services via a single pane of glass is again a minimal requirement. For anyone from a technology background (or those that have been reading my blog posts), there should be no surprises with respect to the NoOps technology stack described above.
Where does Distributed Ledger Technology fit in?
When I look at existing Bioinformatics software and platforms they are exhibit a lot of legacy thinking and technology (with the obvious exception of those that are beginning to exploit IaaS provided by cloud vendors but these tend to be very domain specific solutions).
They are often standalone installations or centralised platforms deployed in private data-centres on physical hardware. They make limited use any kind of PaaS or Container technology (although the sector is beginning to catch up). Furthermore, where transaction logging exists, the transnational data captured is limited and is often stored in a relational database. Identity management is generally also centralised and based on simple role-based access, leaving the organisation that has deployed the software to resolve audit and tractability issues.
It is in the area of transaction authentication, authorisation and attribution where a distributed ledger, integrated as part of the platform, will not only contribute to the resolution of privacy and data sharing issues but may also provide a powerful method for not only fostering innovation but crucially, accelerating the adoption of the platform by the broader community.
An exchange with interchanges not a network with nodes
The BiOMEX platform is a distributed exchange composed of one or more interchanges, each of which defines access to its services via digital contracts. The simplest instance of an interchange can be instantiated on any machine that has access to the internet by retrieving the binary/container from a registry and running the start script. Once and interchange is up and running the bioinformatician will have access to the services listed on the exchange via a standardised API, Web Interface and command line tools.
The precise level of access will be determined by the digital contract associated with the user/interchange. The most complicated exchanges can span multiple hybrid-clouds and consume services offered by other exchanges, but from the perspective of a user, none of this complexity is apparent. The use of a distributed ledger and digital contracts is not a gimmick, it is a fundamental requirement for authentication, authorization and attribution of every transaction (API call) on the BiOMEX.
I am in the process of assembling a team to deliver a working BiOMEX. We already have partners in academia that are extremely excited by the prospect of listing their data and algorithmic assets on the BiOMEX and judging by the general sentiment towards the project I expect many others to follow their lead. Ideally, I’m looking for enthusiastic and motivated engineers, bioinformaticians and technologists to contribute to our enterprise. In return, I’m offering the chance to Bank on Biology and Distributed Ledger Technology.
If you are interested and want to help please contact me directly.