Digital Future of pharmaceutical Industry

  AI in Pharma: The Formula for Success Across the Drug Lifecycle

 

The Indian economy is largely a service-based one and the past few years has only seen that trend intensify with one big exception: the pharmaceutical industry. India has the third largest pharmaceutical industry by value as of 2020. It is growing extremely fast. The Indian pharmaceutical industry started out as copycats (reverse-engineering driven) but then evolved throughout the value chain to become a ‘global leader’ which has been evident from the vaccines development, manufacture on global scale and supply to the world during the pandemic. Indian pharmaceutical industry is very quick to adopt the new technologies in various areas of pharmaceutical value chain. A brief over view of the same is given in this article.

 

ROLE OF DIGITAL TECHNOLOGIES IN DRUG DEVELOPMENT:

 

Getting a new drug to the  market is a long tedious process; it can take many years or even decades. There are all sorts of experiments, clinical studies, and clinical trials that one has to go through. About 90% of all clinical trials in humans fail even after the molecules have been successfully tested in animals. The process is as follows:

Þ     the scientists study  the associations between drugs, diseases, and proteins/targets, and find out what the target for the drug should be, i.e., which protein/target it should bind with;

Þ     formulate what kind of properties are desired from the drug: how soluble it should be, which specific structures it should have to bind with this protein/target, should it treat a particular type of disease;

Þ     brainstorm about which molecules might have these properties (lead molecules); there is a lot to choose from on this stage: e.g. one standard database lists 72 million molecules;

Þ     the  lead molecules, or leads, are actually subjected to experimental validation;

Þ     once the validation of the concept in the lab indicates that the substance works, the whole clinical phase can be initiated -which is a very long and tedious, and only a small percentage of drugs actually go all the way through this funnel and reach the market.

 

So where is the place of digitalization or adaptation of digital technologies (including Machine learning (ML) or Artificial Intelligence (AI)) in this process? Naturally, we can’t hope to replace the lab or, God forbid, clinical trials; we wouldn’t  want to sell a drug unless we are certain that it’s safe and confident that it is effective in a large number of patients. This certainty can only come from actual live experiments. In the future it is likely that we will be able to go from in- silico (in a computer ) to patients immediately with the AL- driven drug discovery pipelines but today we need to study experimentally.

 

Note, however, the initial stage of identifying the lead molecules. At this stage, we cannot be sure of anything, but generation of experimental evidence in the lab is still very slow and expensive, so we would like to find lead molecules as accurately as we can. After all, even if the goal is to treat a particular condition there is no hope to check the entire endless variation of small molecules in the lab. 72 million is just the size of a specific database, the total number of small molecules is estimated to be between …… and ……., and synthesizing and testing a single new molecule in the lab may cost thousands or tens of thousands of dollars. Obviously, the early guessing stage is really, really important.

 

By now one can see how it might be beneficial to apply latest Al techniques to drug discovery. We can use machine learning models to try and choose the molecules that are most likely to have desired properties.

 

But when you have 72 millions of something, “choosing” ceases to look like classification and gets more in to the “generation” part of the spectrum. We have to basically generate a molecule from scratch, and not just some molecule, but a promising candidate for a drug. With modern generative models, we can stop searching for a needle in a haystack and design perfect needles instead.

 

It would take $2.5 billion and about 10 years of research for a drug to be developed from scratch. Only 1 in 10 drugs would pass all necessary  stages and eventually reach the patient. The present-day fast and furious world  can’t  afford such expenses or such time frame. We have practically experienced it after the big blow COVID-19 has given to the world.

 

AI can make the process of drug discovery more efficient by interconnecting different stages in drug discovery which was not feasible so far. Machine learning techniques can effectively interconnect different stages and use the data in the previous stages to understand the subsequent stages. The simultaneous access to this cascading data can be helpful in identifying measurable parts than relying upon non-quantifiable descriptors such as symptoms of a disease etc. it would be possible to map the results of trials conducted on the patients in to molecular signatures thus redefining the disease to the core of its occurrence.

 




AI has been successfully tested in all main stages in drug development
:

Stage 0. Literature overview

Literature review is a step where a scientist would go through lot of research information and  analyze the exact information required for the present. AI can come handy to analyze the relevant data and summarize the required material in a concise form.

 

Stage 1 : Identification of targets for intervention

The first stage in drug development is to identify biological origin of the disease and its resistance mechanisms. To provide antidote it is essential to identity the target proteins/ markers  accurately. The siRNA screening and deep sequencing kind of high-throughput techniques have provided ample amount of data for identifying viable target  pathways. Multiple data sources can be effectively analyzed using Machine learning algorithms to arrive at an appropriate target protein.

 

Stage 2 : Drug candidate discovery

Once the target protein is identified the compound that can interact with the identified target molecule has to be looked for. This process involves the screening of thousands of natural, synthetic and bio-engineered compounds for their effects and side effects. Again, Machine learning can come to the rescue by screening millions of compounds it can throw out best compounds with minimal side effects.

 

Stage 3 : Faster clinical trials

It is very crucial to identify suitable candidates for clinical trials to properly assess the effectiveness of the drug. AI techniques can be harnessed to identify the right candidates for clinical trials and also analyse the results  with accuracy. AI can also play an  advisory role in suggesting the way forward when the results of clinical trials are non-conclusive.

 

Stage 4 : Identification of biomarkers for diagnosing the disease

Biomarkers are the molecules present in the bodily fluids or tissues that Indicate whether the patient has a disease or not. They help in diagnosing the disease and  also to identify the progression of the disease facilitating the doctors to choose the correct treatment methods  and monitor the effect of the drug. AI techniques can help in identifying the appropriate molecule that would function as a  biomarker. Biomarkers can be diagnostic (help to diagnose the disease, Risk (ascertain the risk of patient developing the disease) prognostic (analyse the progression of the disease), or predictive (whether the  patient would respond to that particular drug).

 

The application of AI in drug design and discovery is at its nascent stage. Some pharmaceutical companies like Merck & Co are working on a novel project which employs deep learning to identify novel small molecules. Pfizer has started collaboration with IBM Watson for immuno-oncology drug discovery research.

 

As the discussion “future of AI in medicine” is put forward there would always be two sides for the argument. The general consensus is that the repetitive and quantitative tasks can be delegated to machines while qualitative and cognitive tasks should be done by Humans. After all there is no replacement for human intelligence when it comes to cognitive tasks.

 

TRENDS OF USE OF DIGITAL TECHNOLOGIES IN PHARMACEUTICAL INDUSTRY:

Some of the technologies who have already made in roads in pharmaceutical and healthcare value chain and other promising as well as futuristic technologies are:

 

Cloud-based solutions: Cloud- based databases enable pharmaceutical companies to collect data from multiple stakeholders to build a rich, integrated data repository.

 

Artificial Intelligence: AI can be helpful in accelerating drug discovery and development processes. Some of the startups and pharmaceutical companies using this technology are: Pangaea Data which is a British start up, it uses unsupervised Al to identify cohorts for drug discovery, clinical trials. In Vivo Al, a Canadian startup, develops novel algorithms for drug discovery. Al can help pharmaceutical companies to identify patterns in unstructured data and intuitively manage patient health.

 

Flexible production: The pharmaceutical industry is exploring new ways of manufacturing to cater to the needs of changing market dynamics such as small batches of precision medicine. Single use bioreactors are gaining popularity due its multiple advantages like eliminating the downtime, continuous manufacturing, low energy needs, high productivity and waste minimization, Scottish startup, Cellexus, makes single use bio-reactors which uses patented airlift technology to move nutrients and cells in bubbles.

 

Big data and Analytics:  The pharmaceutical industry requires high performance systems to analyze large volumes of data during drug discovery and development process. Startups like pryml and pomicell work in the area of analyzing large data sets and give customized insights to the clients Analytics can help pharmaceutical companies use the vast amount of digital data available via non- traditional sources like social media and patient forums to supplement traditional data sources  such as physician, enterprise and claims data for detection, prevention and intervention.

 

Additive manufacturing3D printing of tissues and cells has great applications in drug development, organ engineering and regenerative medicine. FabRX is a British startup which makes M3DIMAKER a 3D printer for making personalized pills. Frontier Bio, a US-based startup offers  FLUX-1 a3D bio printer for making human tissues.

 

Block Chain: Data needs to be protected with utmost  priority in pharmaceutical sector to avoid leakage of data regarding drug designs, to avoid counterfeit medicines etc. Pharma Trace, a German startup, offers blockchain – based ecosystem to secure data and deploy smart contracts in pharmaceutical industry. Veratrak is a British startup offering blockchain- based document collaboration and workflow management platform for the pharmaceutical supply chain.

 

Extended Reality: Extended Reality, Mixed Reality (MR), Virtual Reality (VR) and Augmented Reality (AR) have enabled visualizations like never before. US-based startup Name and German based Goodly Innovations help in visualising and understanding the molecular structure of elements and molecules of proteins.

 

Precision medicine: Precision medicine is the idea of customized medicine based on the drug exposure models used to determine the pharmacokinetic and pharmacodynamic properties of drugs. This helps in arriving at a right dosage for drugs based on age, sex, comorbidities and other clinical parameters. Exactcure is a French startup which offers prediction  of the efficacy and interaction of drugs on an individualin real-time. Tepthera, a Swiss startup, offers platform to identify T-cell antigens. The MEDi platform identifies tumour- specific antigens from patient’s human leukocyte antigens.

 


Digitized pills
smart pills with ingestible sensors, smart patches that track and collect patients’ health data, which can be used to intervene and optimise medicine efficacy are the new age pills.

 

Mobile apps: These enable pharmaceutical companies to assist  patients in managing their conditions, in adhering to their medications, and in ‘coaching’ healthy behaviors

 

Digital therapeutics: Digital therapeutics  deliver evidence-based treatment of physical, mental, behavioral conditions of neurocognitive  and neuromotor impairments etc. for example, the VR headset developed by cognitive can help in neurorehabilitation by providing immersive resort world. Dopavision, a German startup, is a smart phone -based digital therapy to control Myopia.

 

Reimagining pharmaceutical industry in the future:

 

The pharmaceutical industry is evolving in response to the evolving behavior of customers, healthcare professionals, and investors, covid-19 has not only reshaped the pharmaceutical industry’s functioning but has also fuelled up development  aspects of the pharmaceutical industry trends- notably in digitalization and patient engagement.

 

The industry is adapting the digital technologies including automation, machine learning, artificial intelligence and data science for improvising the efficacy, effectiveness, speed and cost-effectiveness of the life cycle of the drugs -right from early drug discovery, development, manufacturing, supply chain to the life cycle management. These technologies are helping to build seamless integration of various facets of pharmaceutical and medical sciences such as life sciences, diagnosis, better understanding of diseases, designing and discovering therapeutic interventions and healthcare management thus aligning the pharmaceutical industry’s road map with the shift towards preventive, predictive and personalized therapeutic outcomes.

 

Pharmaceutical innovations  create value to society by making it possible to  generate improvements in patient health (net of treatment risks) that were previously unattainable. It is the uniqueness of such health improvements that defines pharmaceutical innovations.

 

It’s time to embrace what is crystal clear - Digital technologies are rechanging care, the practice of medicine and how pharmaceutical innovations reach patients.

 

 

By,

DR. Shrinivas Savale

CEO, AIC-LMCP Foundation

 

DR. Narayan Darapaneni

Chief Data Scientist

 

Visit the Mundial group page

Comments