Bioinformatics scientist, a doctor, a professor at the Faculty of Medicine, lecturer and philanthropist, and now Machine Learning Tech Lead at HTEC, Ognjen Milicevic is a man of impressive versatility of interests who believes that many exciting branches can unite life sciences and applied mathematics. However, the emerging field of genomics fascinates him the most.
Genomics may be critical to identifying, curing and preventing diseases, and reshaping life sciences industry as we know it. What are the challenges and opportunities ahead? What novel technologies and key talents will reshape the way how genomics augments human lives and experiences in ways previously thought impossible? We caught up with Ognjen to pick his brains and hear his perspectives.
As a young scientist, Ognjen admits he did not have a good plan about his future, but he points out that he has always been versatile but ambitious.
“If I try looking back for patterns, it seems I have always been fascinated with genomics. When in high school I heard somewhere there is a branch called mathematical oncology, for a while I kept telling people this was what I would like to do in future. With all due respect to people in this actual branch (I still believe it is super interesting), genomics is very close to what I imagined at that stage”, explains Ognjen.
He also admits that looking for patterns retrospectively is a seductive practice, something which is frowned upon in medical statistics and data science.
“I would say that the best validation is making predictions – and I predict great things in Life Science domain for us at HTEC.”
Ognjen believes that science is endlessly interdisciplinary and that extreme specializations are a thing of the past. He says that there are many critical components that overlap to build a certain discipline, which also applies to genomics.
“Two years ago, during the pandemic, I was part of a genomic initiative for a COVID-related sequencing, and everybody was an impressive expert in their field. We had some funds available and wanted to design a cutting-edge project from scratch. It starts like a joke – a virologist, immunologist, and an engineer walk into a bar and order a project. You need to be interdisciplinary trained to even think about it”, says Ognjen.
He explains that this is not only managerial stuff.
“As a genomics scientist, you are positioned at the computer between the lab and the clinician. You choose the algorithm. You validate what came before, and make sure the result makes sense. You enable the people after you. You must keep yourself multi-faceted to fit in that unusually shaped void. The good thing is that this growth comes naturally for people in love with this field – there is no better place to learn.”
Almost thirty years ago, The Human Genome Project created over 20 university labs, which resulted in a decade of work and billions of dollars globally. As a result, now we have gene therapy drugs (like SMA), we managed to create STEM cells from ordinary human cells, we improved cancer therapy and understanding, introduced minimally invasive gene-based pregnancy testing, etc.
However, today, genome sequencing time and effort has been reduced to a few hundred bucks. It turned out that technology was faster than medicine. The promise of using big data to unlock the full potential of genomics is exciting, but there are challenges ahead.
“Scientists are excitable by nature due to their passion, and human genome project was romanticized as the “Code of God” that would propel us into a new era of personalized medicine. The expectation is still strong, but the hype subsided in the medical community when our genome turned out to be like Pythia – it gives cryptic answers that could go either way. Commercially, it is very exploitable, but the conclusions are still not up to our rigorous clinical standards. You need human genomics scientists to interpret the answers. You need teams and collaboration.”
Source: NCBI Insights
Ognjen explains that genomics is currently used as like any old-school procedure yielding simple data where we have this averaged information with no actionable points.
“You notice a point mutation that is associated with a risk of getting a disease, and then you check for that mutation in your patient and tell them they are on average under greater risk. This, in a way, resembles measuring the belly circumference of a patient and telling them they are under risk for diabetes; this is sufficient as it justifies a change in lifestyle which is bound to be good, and we don’t really need to know which patient would ultimately suffer from diabetes – losing weight would benefit each patient.”
He emphasises that to truly take advantage of this information, we need to understand the interplay across the genome, and opt for personalized medicine.
“I would say the opportunities and expectations of genomics are both the same – to go from the individual letters to the story written in our genes. And, currently, we are only succeeding in reading some words and an occasional sentence. This is not a poetic analogy — it is a realistic reflection of the complexity that goes way beyond the simple DNA code.”
Genomics has several specific pipelining standards akin to programming languages and IT frameworks, but this is only a small proportion of the skills needed. Ognjen points out that most of the requirements are divided between Cloud technologies, high-performance computing, quality assurance and Data Science.
As an innovation enabler and backed up by many years of experience working on innovative MedTech solutions, HTEC has an engineering excellence, strong know-how and deep expertise in this domain to bring life-changing solutions to life.
“I feel that our DevOps are more than equipped to handle the infrastructure, our QAs can take on Big Data, while our data scientists are shaped to be capable of exploring the domain of clients’ interests and adapt the statistical tools to the informational content. Since every case is slightly different, computational optimization and research are daily tasks, and these are performed better in a culture promoting self-growth. Finally, all of these steps are overarched by security, which is easily overlooked but critical for all other work segments.”
In Ognjen’s view, HTEC distinguishes from the rest on the market with its ability to create the environment of self-growth and strong passion for choosing noble projects and industries over the low-hanging fruit like gambling.
“HTEC started off with projects like Humeds, and it shaped us as a collective. People regularly email me from within the company to say that they would love to be involved in future MedTech projects. With such passion and good technical basis, we can tackle any medical project including ones from genomics.”
20 years ago, organisations were starving for data. Now the data is catching up. Ognjen believes that the synergy of Academia and the industry can help unlock the full value of data and utilize it for worthwhile research and commercial apps.
“The data is catching up, but whether you have access to it is a whole different matter. Although we have experience as a company in working with sensitive medical data belonging to the client, there are many fantastic datasets to augment your models with. Academia has the access rights, and thus contains the expertise to process that. Our close ties with Academia complement the commercial client-oriented experience to create a consistent timeline of improvements, at the personal and company levels.”
Implementing a scalable, powerful and searchable genomics-focused big data ecosystem will pave the way for futuristic commercial and academic advances worldwide.
“Simply put — academia and industry complement each other in full, and I believe we have a very deep understanding of the values this symbiosis can bring. However, even the most organized projects still need data scientists to ensure that all data and metadata complies with FHIR standards.”
Source: analyticalscience.wiley.com
The most private data is human genome. While the developments described above will massively contribute to our understanding of the human genome and how it is related to human disease and other various important issues, they also pose new risks in terms of privacy related harms. Regulatory compliance plays a critical role here.
“Compliance is two-fold: you have the security regulation that governs the day-to-day activities, but we must also keep in mind that translation to clinical practise is the ultimate goal. All the successful projects will, at some point, go through regulatory certification, and the compliance steers you towards it and navigates you through it. Without proper compliance in genomics, some tasks are later more difficult, while others are completely impossible.”
The regulation can still be slow to catch up, so using common sense is not enough. Respecting country boundaries is often a requirement, especially for national genome projects that are happening all around the world. This has implications for execution, monitoring, interpretation and so many other work aspects, and not pursuing regulatory compliance with the purpose will cause all companies’ work plans and predictions completely to fail.
“Luckily, we have great experts in the company for compliance, and genomics is regarded as Software as a medical device. The company already went through the entire certification procedure including our own product HUMEDS, and we feel confident working on novel genomic projects for our clients repeating hand-in-hand with our clients.”
—
To learn more about how we can help you unlock the data and reshape the future of genomics to contribute to better care for all, talk to our experts or connect with Ognjen Milicevic. In the meantime, check out our success stories to see how we help provide better health for all.
