If modern drug discovery is a game of needles and haystacks, researchers need every bit of help they can get to understand the underlying mechanisms of disease and bring the right therapy to bear against the right target. Biomarkers – alone or in combination with others – are objectively measurable characteristics that can indicate the presence or risk of disease-causing pathogenic processes, or predict the physiological response to a particular treatment.

As such, identifying new biomarkers has become an increasing focus in 21st century drug discovery and development, and the inclusion of molecular markers in clinical trial design allows for the retrospective analysis of data and the identification of patient sub-groups that are likely to respond to a given therapy.

As for technology platforms to identify biomarkers that inform drug discovery efforts, the systems and techniques involved are intricate lab processes, from tissue microarrays and gene expression profiling to proteomics and mass spectrometry, as well as machine learning and other bioinformatics-based methods of interpreting data and validating whether biomarkers have diagnostic, predictive or prognostic value. The range of technologies being employed to identify new biomarkers is vast, so below is an exploration of promising case studies and talking points from different corners of the medical industry.

Epigenetics: biomarkers based on gene expression

Part of the legacy of the Human Genome Project and its mapping of all three billion base pairs in the human genome is the boon it has provided for biomarker research. Technologies developed as part of sequencing the human genome were adapted for use in molecular diagnostics and the discovery of biomarkers, helping to drive the development of gene expression analysis.

Epigenetics – the mechanisms by which gene expression and activity is heritably altered without changing the underlying DNA sequence – represents a further step towards identifying new biomarkers that, through the study of epigenetic deregulation, lend insight into the progression of human disease.

As Kewal K. Jain puts it in ‘The Handbook of Biomarkers’: “Whereas HGP [Human Genome Project] provides the blueprint for life, the HEP [Human Epigenome Project] tells us how this whole thing gets executed, what determines when and where genes are switched on and off to produce a person. And knowing more about the human epigenome may provide clues to what goes wrong in cancer and other diseases.”

How well do you really know your competitors?

Access the most comprehensive Company Profiles on the market, powered by GlobalData. Save hours of research. Gain competitive edge.

Company Profile – free sample

Thank you!

Your download email will arrive shortly

Not ready to buy yet? Download a free sample

We are confident about the unique quality of our Company Profiles. However, we want you to make the most beneficial decision for your business, so we offer a free sample that you can download by submitting the below form

By GlobalData
Visit our Privacy Policy for more information about our services, how we may use, process and share your personal data, including information of your rights in respect of your personal data and how you can unsubscribe from future marketing communications. Our services are intended for corporate subscribers and you warrant that the email address submitted is your corporate email address.

A good example of a technology leveraging epigenetics to find new biomarkers is Oxford BioDynamics’ (OBD) proprietary industrial platform, EpiSwitch. The high-throughput platform is designed for rapid identification of chromosome conformation signatures (CCSs), combinations of DNA sequences across the genome that modulate gene expression.

EpiSwitch is already in demand to generate potential new biomarkers of various types, from predictive biomarkers tracking responses to immuno-oncology treatments to prognostic biomarkers in lymphoma patients. Most recently, OBD signed a collaboration agreement with Casa Sollievo della Sofferenza, a renowned biomedical research institute in southern Italy, to use EpiSwitch to identify new diagnostic biomarkers for autism spectrum disorder (ASD).

Under the terms of the agreement, which was announced in November, Casa Sollievo will supply blood samples from autistic and non-autistic patients, which will be screened at high resolution for CCSs by EpiSwitch. The partners’ aim is to use identified epigenetic biomarkers to develop a blood-based diagnostic assay for ASD. With incidence rates of autism skyrocketing in many countries and the prospect of more effective interventions with earlier diagnosis, the project has the potential to improve quality of life for millions of ASD patients, as well as expanding the epigenetic knowledge base for the neurodevelopmental condition.

“Diagnosis for ASD remains challenging due to broad-ranging symptoms and differential progression and manifestation of the disorder,” said OBD in a press release. “As the EpiSwitch platform monitors the environmental impact on the genome, it is well placed to identify biomarkers for diagnosing ASD through minimally invasive blood sampling.”

Vocal biomarkers

While biomarkers are more commonly associated with biological properties or molecules measured in blood or tissue, the human body is a data-generating powerhouse and, with the right technologies, potentially relevant biomarkers can be gathered from some surprising places. In the last few years, interest has been building around the identification of diagnostic biomarkers in the human voice; in years to come, the nuances of our voices might provide as much insight into our health as a round of blood tests.

A growing community of tech firms has formed around the idea of vocal biomarkers, from computing behemoth IBM to smaller start-ups and voice analysis specialists including Sonde Health, Beyond Verbal and Cogito Corporation. Major partnerships have boosted the profile of the field; in 2016 the US Army Medical Materiel Agency (USAMMA) partnered with researchers at MIT Lincoln Laboratory to work on a computer algorithm to help diagnose traumatic brain injury based on vocal biomarkers such as voice fluctuation, the elongation of syllables and the coordination of tongue and lip movement.

“The ultimate goal is a US Food and Drug Administration-cleared, real-time mild TBI screening app and hardware device which can be used throughout the echelons of care from point of injury to rehabilitation,” said USAMMA product manager Brian Dacanay.

More recently, Boston-based Sonde Health announced last year that it had secured new patents in the US and Australia covering its technology for analysing vocal biomarkers for the screening of both psychological and physical disorders. The firm’s voice analysis platform, based on technology licenced from Lincoln Laboratory, has produced promising results in pilot studies measuring biomarkers for the likes of Parkinson’s disease, depression and cognitive impairment, with potential to also identify diagnostic biomarkers in cardiovascular and respiratory disease.

“[The human voice] requires incredible coordination of multiple brain circuits, large areas of the brain, coordinated very closely with the musculoskeletal system…it’s also involved with the respiratory system, which is required to activate the vocal chords,” Sonde Health co-founder Dr Jim Harper told MobiHealthNews. What we’ve learned is that changes in the physiological state associated with each of these systems can be reflected in measurable, objective features that are acoustics in the voice.”

AI and machine learning: vital but imperfect

The data-crunching requirements of the vocal biomarker field – the development of which is contingent on building up huge banks of voice data and having the means to effectively interpret it – brings to mind the immense importance of artificial intelligence and machine learning systems when trying to map biomarker patterns and draw conclusions about disease development.

Machine learning and neural networks represent a much-hyped facilitator of biomarker discovery, and one that has sometimes been presented as a cure-all for researchers’ data woes. The promise of machine learning is the ability to take a more empirical approach to biomarker identification, allowing the data to speak for itself without having to start with a specific hypothesis to test.

It’s certainly true that machine learning models open up a new world of statistical analysis and have the ability to spot anomalies and identify potential biomarkers that human researcher would miss, but with datasets and objectives as complex as those in biomarker discovery, it’s important not to lean too hard on AI-derived results without well-considered oversight and validation processes, as false positives are a strong possibility when machine learning technologies make calculations based on limited sample sizes and with no defined route to their statistical objectives.

“Statistical methods will always take the easiest way out to the answer you ‘want’,” warned data scientist Imran Haque in a recent interview with Forbes. “Right for the wrong reasons looks right in your small data set, even if it is stops working in a larger set.”

As the frontiers of biomarker discovery continue to open up, medical devices and technologies will play a vital role in the collection, analysis and validation of previously hidden markers that will unlock new insights into some of the world’s most persistent health issues. Bringing it all together in service of patients will require a great deal of collaboration between disciplines, from clinical researchers and data scientists to machine learning experts and medical technology manufacturers, but the pay-off is great enough that every effort must be made to make the best possible use of technology in the hunt for new biomarkers.