2.5 quintillion bytes of data are created every day. The challenge is to identify the relevant ones for solving a challenging question. That is where Artificial Intelligence solutions and blockchain can help make this knowledge public.
If a beloved family member or close friend becomes seriously ill, you want the best treatment and cure for them. This is how the idea of Innoplexus came about. "A few years ago, a good friend and mentor of Dr. Gunjan Bhardwaj, Founder and Global CEO Innoplexus, was diagnosed with cancer. That's when we realised how limited the opportunities are for individuals to access sound and understandable medical knowledge. So, as a patient or relative, you feel helpless when you want to address alternative therapies or consolidate a second opinion based on research areas or therapeutic success,“ recalls Kumar Anshu, CEO Innoplexus, India. Out of this crisis, a business idea to create a digital system that answers medical questions and democratises information was born. But this was a challenge, since organisations work in silos – both internally and externally.
Innoplexus came up with a solution that could facilitate better ways of discovering, exploring and analysing biomedical research. Their AI-based core product, iPlexusTM, is an end-to-end platform that generates intelligence and insights spanning pre-clinical, clinical, regulatory and commercial stages of a drug, across therapeutic areas and indications. It is an innovative and intuitive semantic search facility which understands biomedical concepts and allows users to see relevant results in context across multiple data sources like publications, clinical trials, congresses, theses and more.
It offers users quick, crisp and summarised snapshots of the current landscape in any given context, be it important deals and mergers, fast-tracked clinical trials and much more, helping them make informed and data-backed decisions. iPlexusTM has more than 300 terabytes of crawled and indexed scientific data across 365k clinical trial databases, 200 biological databases, all major patent offices, regulatory agencies, patient forums, which includes 25 million publications, 20 million patents. This number is continuously growing. An analysis layer applies system intelligence and machine learning to make sense of the information retrieved (deriving inferences), leading to more related and relevant possibilities.
A lot of research that takes place at small pharma labs, niche biotech companies, or academic institutes etc. is lost because it is scattered, unstructured and not readily available through a common point of source. Innoplexus aims to bridge this gap by collecting, cleaning, categorizing and arranging various kinds of information ready for the user.
“We built iPlexusTM as a discovery engine – and created a deep, dense and diverse real-time data ocean, using Artificial Intelligence to crawl, analyse and aggregate structured and unstructured public data. Since our experts trained our AI in the specific language used in life sciences, our customers get relevant insights through intuitive, interactive visualization. We constructed a relevant scientific library,” says Anshu. “And we still know there is information missing.”
History, as it is often said, is written by the winner. The history of research is no different, and it is rare for failures to be lauded in science. Non-governmental organisations, contractors and researchers want a good track record, those funding the research need to show that they are spending wisely, and journal editors want to highlight breakthroughs.
Publication bias affects the body of scientific knowledge towards positive results. This means that the results of thousands of experiments that fail to confirm the efficacy of a treatment or vaccine – including the outcomes of clinical trials – fail to see the light of day. So organisations will have no idea if a particular experiment has already been conducted in an “unpublished” domain, especially if it comes to failed experiments.
Despite their potential, negative results are repeatedly relegated to the lab books, the drawers and the trash bins. This is not a new phenomenon – research published in Controlled Clinical Trials in 1987 showed that statistically significant clinical trial results were three times more likely to be published than those supporting the null hypothesis.
In 2015, the World Health Organisation (WHO) announced its position on publishing the results of clinical trials, calling for the main findings to be submitted for publication in a peer reviewed journal within 12 months of study completion, and for all previously unreported results – including negative findings – to be published.
Initiatives like the All Trials campaign are trying to balance the literature by encouraging researchers to publish their negative results. The best available evidence shows that around half of all the clinical trials that have ever been carried out have never reported results. It means that information about the medicines we use every day is at risk of being lost forever. But it is not just the clinical trials literature that is missing negative results – any experimental discipline that works on the basis of a hypothesis runs the risk of this bias.
In Disease Models & Mechanisms, Dr. Natalie Matosin and her colleagues from the University of Wollongong and the Schizophrenia Research Institute in Sydney, Australia claim that “Science is, by its nature, a collaborative discipline, and one of the principle reasons why we should report negative results is, so that our colleagues do not waste their time and resources repeating our findings.” The basic principle of science is to get things right by analyzing what went wrong.
Results of preclinical studies and their raw data are regularly published with a delay of 18 to 24 months or never. Often, similar studies are being carried out elsewhere, slowing and increasing the cost of developing new medicines and therapies. The reasons are concerns about priority of publication and data sovereignty, the duration of peer reviews and a lack of incentive. There is presently no formal, safe and mutually beneficial way to share unpublished life science results. Currently, there is no known platform which helps organizations purchase license to work done by researchers.
Innoplexus solves exactly this problem. “We help researchers publish their “unpublished” work (experiment data) on our blockchain based platform; and enable purchase of license to use the “unpublished” data by other organizations, thereby saving time and cost. The iPlexusTM blockchain-based publishing platform will enable a single view that includes published data, unpublished data, restricted data, enterprise data and third-party data”, explains Abhijit Keskar, Vice President Technology. “We add this to our knowledge and function as an innovative collaboration tool.”
Abhijit describes their newest version of iPlexusTM Innoplexus stretch technology: “Our platform brings together potential demand and providers of raw data and results from preclinical studies. Blockchain technology and included Smart Contracts will unalterably secure the authorship of the data and its authorized users. Our AI determines the value of the data by means of their relevance and the reputation of the providers. For the research institute this can be an additional source of income. And for the individual researcher there is an incentive to offer his data from the documentation of the time of emergence, which is important for patent issues.”
In simple terms, blockchain is a database of cryptographically secured information that is replicated and distributed across a network of independent, decentralized nodes. Typical attributes of a blockchain are that it is trusted, shared, secure, traceable, and a single source of truth. The use of predefined smart contracts would not only make sure that trust is maintained throughout the process but will also ensure that the records on blockchain will act as a valid IT proof in case of any IP dispute.
iPlexusTM is a network platform for pharma companies and authors of experiments, who are interested in publishing and trading the license. Innoplexus is the intermediary and helps users discover what other work the author has done in the public domain, thereby enabling well-informed decisions. The Innoplexus model is based on Blockchain and AI. Blockchain is used for three main reasons:
The patent-pending document valuation engine is based on AI and is unbiased as it takes into consideration earlier valuations as well. Valuation is “the” parameter that reflects quality of the experiment and its utility to the specific user. Since machine learning is trained using the inputs from experts in the field, it is a fair valuation that reflects quality of the experiment. Thereby, the experimental documents are never passed to the platform.
With their blockchain technology, Innoplexus want to accelerate the development of drugs reducing the cost of development based on a broader knowledge base, strengthening healthcare for humans and animals through practical innovation acceleration. “Thanks to its proprietary CAAVTM (crawling, aggregating, analysing, visualizing) based platform, iPlexusTM provides the entire Life Science knowledge along with daily updates. Therefore, any unpublished data can be validated in real time with the world's entire published work, thereby eliminating any plagiarism or double publication. That’s really unique,” claims Abhijit proudly.
“We are aware that the new iPlexusTM combining AI and blockchain in order to democratize research data may need to be defended against protectionist aspirations, and that it is important to promote the use of the platform, in Europe and the US”, admits Anshu. “iPlexusTM is a one-stop solution for all life sciences data – structured, unstructured, published and unpublished - which enables real-time actionable insights and to help create a previously unrealized resource for the industry. And we are enabling research data to be made available and demanded digitization of data to avoid loss of efficiency and knowledge.”
Innoplexus wants to open data silos and increase the number of publications with its own research publishing network based on blockchain. Researchers will be able to publish their unpublished work and get a validation of their work. Organizations will collaborate and avoid re-invention of wheel. This will result in faster launch of new drugs, thereby saving lives.