Deep phenotyping brings accuracy to precision medicine

Biotech firms are leveraging the wealth of omics and imaging information to examine deep phenotypes for a spread of scientific purposes together with precision medicine approaches to liver illness, neurodegenerative problems and infertility.

With dozens of omics and imaging applied sciences in its arsenal, the biopharma trade is producing unprecedented quantities of knowledge in contrast to just some years in the past. And the standard of this information is best too. For instance, researchers have made progress in reliably studying hard-to-sequence areas of the human genome and exploring uncharted proteins within the so-called darkish proteome. 

The rising richness and complexity of organic information are what present the depth in deep phenotyping, an rising pattern catalyzing thrilling developments in customized and precision medicine. In the context of healthcare, the time period deep phenotyping usually refers to the excellent evaluation of illness traits, or phenotypes, as an end result of genetic, epigenetic, way of life or environmental elements.

Precision medicine facilities on the person, stressing the truth that every affected person is totally different. However, precision doesn’t assure accuracy in mapping medical interventions to outcomes. This is as a result of precision medicine doesn’t account for the extent of complexity that underlies ailments and, in consequence, fails to set up the causality of illness, as an alternative specializing in one or few biomarkers which are vaguely correlated to illness development. 

Deep phenotyping brings accuracy to precision medicine by embracing the complicated nature of organic techniques and learning phenotypes at a number of ranges by combining a number of approaches. This is already making an affect on quite a few therapeutic areas together with liver illness, neurodegenerative problems and infertility.

Better therapies for liver illness

Advances in transplant medicine and a greater understanding of liver biology haven’t translated into a major change in liver transplant outcomes. As individuals now stay longer lives and have fattier diets, the standard of livers accessible for transplants has worsened. British biotech Ochre Bio goals to enhance the standard of transplant livers with deep phenotyping. Last 12 months, it raised $9.6 million in seed funding to advance its work on precision RNA therapeutics that it’ll use to rework suboptimal livers. 

The startup combines computation, automation and deep phenotyping to develop medicine for liver ailments. Quin Wills, co-founder and CSO at Ochre Bio, defined that the corporate “photographs every liver at the very least six instances and makes use of machine studying to extract options of histology.” 

“We map these histological options to the scientific phenotype of the sufferers that we examine,” Wills added. “At the following stage down, we examine all of the liver enzymes which are circulating by means of the liver.”


By observing how the phenotypes change over time, Ochre Bio’s deep phenotyping method creates higher computational fashions of liver biology. This offers wealthy insights into how the liver reacts in response to totally different medicine. 

With deep phenotyping, the target isn’t about one form of phenotype however about integrating data from totally different organic ranges starting from proteins and different metabolites to full organs. Instead of learning how a therapeutic alters the degrees of a single biomarker or impacts a scientific endpoint, it seeks to reveal the holistic modifications throughout many ranges in any system.

For Ochre Bio, “main phenotypes are the translatable endpoints that you really want to change in a scientific trial whereas secondary phenotypes are properly understood and validated biomarkers,” Wills stated. “The tertiary phenotypes are actually all the small print of mechanistic research utilizing single-cell sequencing.”

The consideration is now more and more on ailments which have evaded therapies up to now and for which growing new therapeutics is costlier regardless of advances in drug improvement. Leveraging the explosion of knowledge and higher instruments to analyze it, deep phenotyping permits firms to examine illness states in additional element than ever earlier than, revealing new mechanisms to goal with novel medicine.

Accurate organoid fashions

Advances in machine learning-based picture evaluation and computation enable biologists to analyze bigger quantities of knowledge and acquire richer insights than what was potential a decade in the past.

For occasion, OrganoTherapeutics, a startup based mostly in Luxembourg, is learning mind phenotypes of sufferers with Parkinson’s illness. Among different areas within the mind, Parkinson’s illness impacts the midbrain, the half liable for motor exercise. Studying the modifications within the midbrain is vital to answering vital questions on how the illness progresses and its affect on sufferers. 

OrganoTherapeutics creates patient-derived midbrain organoids that recapitulate the pathology of the sufferers. Jens Schwamborn, the startup’s co-founder, stated that deep phenotyping permits the agency to “actually take a look at the organoids in a complete method at loads of options on the identical time. They [the researchers] put them collectively, attempt to perceive them, visualize them and see what is definitely totally different in a affected person compared to the wholesome management.” 

OrganoTherapeutics makes use of picture recognition and picture evaluation algorithms to extract totally different options from photographs of midbrain organoids. These vary from easy measurements like what number of cells there are and if they’re useless or alive to how complicated the neurons are when it comes to the variety of branches and nodes, amongst different points. 

“The attention-grabbing half then is that historically we might have checked out one characteristic at a time and in contrast the way it differs between the wholesome versus diseased cells, if in any respect,” Schwamborn added. “But now we’re in a position to take a look at all these options holistically on the identical time.” The startup just lately collaborated with U.S. biotech Vyant Bio to develop clinically translatable assays for Parkinson’s. 


Systems biology emphasizes that at the very least among the properties of any system (on this case a diseased state) come up from the interactions of its elements. As Schwamborn clarified, “there could possibly be 10 small variations that individually in all probability don’t matter an excessive amount of. But when you take a look at all of them collectively, it provides a really totally different image and tells you which of them processes are literally affected. And then lastly, we combine these information with these from transcriptomics, metabolomics or mitochondrial exercise assays to actually take a look at an entire image.”

Systems biology has lengthy been dominated by the event of ever quicker, cheaper and better throughput instruments to examine totally different sorts of biomolecules and cells. Consequently, biologists now generate massive quantities of knowledge. Deep phenotyping has the potential to speed up the interpretation of those information into scientific insights for correct, customized medicine. 

Boosting fertility therapies

When talking of affected person outcomes, biology or expertise aren’t all the time the one figuring out elements. For many dad and mom making an attempt to conceive by means of in vitro fertilization (IVF), the variety of possibilities that they get is proscribed by socio-economic elements, making it an emotionally laborious selection. Positive affected person outcomes — IVF procedures main to pregnancies — rely closely on the flexibility of predictive fashions to precisely assess the percentages of success. For many, the percentages could also be so low as to not even justify the prices.

Traditional prediction fashions for fertility therapy place loads of weight on age. However, age solely accounts for about 50% of the prediction. Consequently, these fashions are inaccurate for almost all of sufferers. The U.S. firm Univfy combines machine studying and deep phenotyping to combine a number of elements in its prediction fashions, thereby bettering the prediction of stay start outcomes for IVF.

A scientific analysis could inform a affected person why she has bother conceiving, however, in accordance to Mylene Yao, co-founder and CEO of Univfy, “it alone will not be useful sufficient for the affected person to decide about therapy. That’s the rationale why we’re deep phenotyping as a broader manner to classify scientific traits and scientific profiles.”

Univfy has established that there are various scientific elements that can be utilized as predictors. “If you apply machine studying to construct prediction fashions based mostly on these elements, they are often very correct,” Yao added. “As a consequence, medical doctors can present a personalised and correct prognosis for these sufferers.”

In December 2021, Univfy raised $6 million in Series B funding. The firm plans to use it to broaden the variety of fertility facilities it really works with in addition to working with employers that supply fertility or family-building advantages. Furthermore, it’s “amassing extra high-quality and complete fertility information and outcomes,” Yao commented. “We’re going to have the opportunity to construct further prediction fashions that particularly deal with every sort of affected person section.” 

Despite advances in biopharma and medicine, drug builders have struggled to sort out many complicated ailments that affect hundreds of thousands globally. Nevertheless, researchers now have entry to larger and richer information on totally different points of human biology. Deep phenotyping will likely be vital to changing these information into insights that meet the potential of precision medicine and enhance affected person outcomes for treatable circumstances.

Recommended For You