We believe that three trends currently characterize the healthcare sector:
- Healthcare industry: The regulatory challenges are increasing while most of the easy pathologies have been addressed. This results in two facts: the R&D investment in the industry is exponentially increasing, and the return on investment is decreasing. On top of this, the health economics requirements are being raised. Public health organizations are more and more reluctant to reimburse therapies a priori, switching to a pay-for-performance setting.
- The empowered patient: More and more patients want to be involved in the management of their own health. More than half of patients get information about their health online before even getting advice from their physician. People also expect to be cured with tailored therapies, with a high probability of success for themselves through a better drug selection/recommendation.
- Data deluge: Massive amounts of data are generated from both the aforementioned parties. Patients publish data everywhere through social networks, to the extent that medical social networks have been developed (PatientsLikeMe, CureTogether, Esperity …). Public health entities are also sitting on huge databases that are about to be made somehow accessible. On the other hand, biotech and pharma companies generate more and more clinical, imaging and various types of -omics data (genomics, proteomics, matabolomics, epigenomics,…). There is therefore a need for powerful tools able to deal with big data.
Vision & Mission
In this context, DNAlytics aims at being the reference data-related solution provider for personalized and predictive medicine, from R&D to patients. In particular, DNAlytics will perform data analysis at various levels of clients’ development pipelines in order to design innovative diagnostic, prognostic and theranostic1 (the field of treatment response prediction or adverse event prediction) solutions. These analyses will also cover automated biomarker identification. DNAlytics will furthermore provide IT tools to deploy those predictive solutions, over the web, directly to clinicians and/or patients, in full compliance with the highest quality standards and regulations.
The resulting benefits of this vision are fourfold:
- for the healthcare industry: reductions of development risk and duration, increased ROI, as well as opening new avenues of development.
- for the diagnostic industry: provide real diagnostics and decision support tools, instead of merely providing measurement technologies.
- for the patient: a better and/or extended life through more efficient and tailored health management.
- for society: reduced public health expenses (since therapies will get more targeted and effective, and also because the diagnosis will be posed earlier and more accurately)
To reach its objectives, DNAlytics has set up a technology platform providing expertise in data mining, machine learning, statistics, intensive computing data management and web technologies. This set of expertise is complementary to those usually encountered in the healthcare industry. DNAlytics uses this technology platform in two complementary ways:
- Supporting its biotech/pharma/IVD clients developing new personalized and predictive medicine solutions
- Developing its own personalized and predictive medicine products, deployed through an innovative IT platform.
The development of new personalized medicine solutions mostly covers the cases of Diagnostic, Prognostic or Theranostic applications. Most of the therapeutic areas today constituting a challenge require the implementation of multimarker strategies. Data analysis and/or modeling challenges are generally involved in these ambitious projects, even at the stage of sample sizing clinical studies. Once the research programs are adequately designed, and the data collected, how does one deal with heterogeneous data combination, given that nowadays, research programs generally combine some -omics technology, clinical records and, for example, imaging technology?
The modeling part has to be correctly addressed so that the diagnostic, prognostic, theranostic models and associated multivariate marker signatures that have been designed are not only satisfactory on the cohorts at hand, but most of all on new, unseen patients. Even after the R&D phase, extra issues are to be overcome. First, how to cross the bridge between a clinically validated multimarker solution and an IVD product available in routine? There is a need for an intermediate software layer between the simple “measurement” of the markers and the production of a decision support value for the clinicians and/or the patients. Second, the market access and reimbursement dossiers will necessarily be different in the case of a multimarker solution combined with a mathematical model than in a traditional case.
Unlike many fields of expertise, data mining is generally not present in traditional life sciences organizations, and this is precisely why DNAlytics will help.