Genomics generates massive datasets that are used to discover, research and develop of new therapeutics globally. It won’t be hard to believe that 3 billion base pairs that form the human genome can now be analyzed through artificial intelligence to find genetic variations in population.
By 2026, big names in pharma such as Astra Zeneca are planning to analyze up to 2 million genomes and study huge amounts of patient data points from their drug clinical trials. That’s the power of AI in genomics.
As we adopt more tools, AI in genomics can be used for different omics studies e.g. transcriptomics. Currently healthcare companies are using AI in combination with HEOR (Health economics outcome research) i.e. AI is being used is to integrate data generated from genomic analyses with research from scientific literature to help find potential clinically relevant genes.
How is artificial intelligence & machine learning used in genomics?
Two main artificial intelligence and machine learning applications in genomics are: identification and treatment. State-of-the-art technology is used to identify harmful genes as well as treat genetic diseases.
Identification of harmful genes
It is extremely tiring and time-consuming to manually analyze huge amounts of data that is present in an individual’s DNA and extract valuable insights from it. Data analytics, powered by AI, can make this process more efficient, accurate and fast, helping healthcare companies and workers make crucial decisions and predictions.
Advanced machine learning algorithms have the capability to compare various gene expression levels in normal and malignant tissue samples of a cancer patient, based on which, predictions can be made to discover which genes are mutated in the patient’s DNA. The ML algorithm learns and provides predications based on the how often a gene is expressed in a malignant sample and compare it with a normal sample with the same gene, additionally infusing new information with every new set of data that is provided.
Artificial intelligence is also used for the identification of genetic mutations within tumours with 3D imaging. For example: An upcoming technology could identify “glioma” type of tumor that starts in the glial cells of the brain or the spine using brain scans of a patient with high accuracy. Using technologies that are based on deep learning and neural networks, treatment process can become highly enhanced wherein doctors won’t require tissue samples to be collected from a biopsy and can rule out the risks associated with surgery. AI & ML provide a vast horizon of possibilities with automation of diagnosis.
These new approaches are only the initial steps in our shift from generalized medicine (which is the current system based on population averages) to precision medicine (which is based on the biology of individual patient) to predictive medicine (which is based on AI-generated insights about an individual’s future health state).
This significant shift in healthcare will ensure that people have their genomes sequenced which will be the foundation of their treatment. Big data analytics can then be further used to compare a patient’s genotypes (what genes convey) to their phenotypes (how genes are expressed in their lifetime).
Treatment of genetic diseases
A fundamental aspect of gene editing is its ability to eliminate the disease-causing genes. While technologies such as CRISPR-Cas9, a genome editing technology which stands for Clustered Regularly Interspaced Short Palindromic Repeats, is used to edit DNA at precise locations with systems that are programmed to target specific stretches of genetic code. With these systems, it is possible to permanently modify genes and, in the future, correct gene mutations are accurate locations in order to treat genetic diseases.
Recently, Emmanuelle Charpentier and Jennifer A. Doudna were awarded the Nobel Prize in chemistry 2020 for the development of a method for genome editing. They have discovered one of gene technology’s state-of-the-art tools – The CRISPR/Cas9 genetic scissors. With this tool it is possible to change the DNA of microorganisms, plants and animals with remarkably high precision.
Machine learning algorithms are used to identify the precise location for DNA alteration. It also provides insights about how to ensure that the repair process of the DNA strand is successful, helping in reducing potential mistakes during the entire process. DNA repair after the process of gene editing is another high potential area where AI can be useful. Researchers would be able to computationally predict the right guide RNAs that would reproduce exact mutations, which will help them develop better research models for studying genetic diseases.
Read more: #10 AI Innovations Disrupting Healthcare
Artificial intelligence in gene technology
Gene editing
Technologies such as CRISPR-Cas9 can edit the DNA sequences to correct defects in genes and treat diseases. Although the technology is extremely precise in targeting the accurate location, there is a probability of mutations due to off-target editing.
Genome sequencing
Machine learning algorithms accelerate the analysis of sequenced data while predicting the genetic alterations associated with a specific genetic disease. This helps to cut down the effort and time needed to develop a medicine.
Predictive Genetic Testing & Preventive Medicine
Genetic screening of new-borns is increasingly becoming a standard practice. This non-invasive genetic screening can detect diseases like Down syndrome during pregnancy. Artificial intelligence can predict outcomes and the risks associated in curing genetic diseases, based on available data.
Challenges and limitations
The fundamental nature of artificial intelligence is to mimic human intelligence or maybe even more -exhibit superhuman intelligence, can interpret complex data and extract valuable information. While AI has immense power, it can also at times be unethical , and make discriminatory conclusions when applied in healthcare. Although AI reduces technical errors of gene editing and mutation identification to a great extent, it also is recognized for enhancing the safety of the procedure. However, several ethical questions remain. People have raised concerns regarding its malfunctioning and believe that it is more dangerous than useful.
Ethical questions such as inequality to access the gene technology based on wealth, using genome editing for purposes other than curing genetic diseases such as enhancement of physical features etc remain unanswered. There could also be several religious objections. Not to forget that the accuracy and unbiased nature of machine learning algorithms is dependent upon the quality of data fed and the developers who develop the algorithm.
Insights into gene editing and gene therapy market
Gene editing market
The gene editing industry has witnessed an expansion due to the rising prevalence of genetic disorders and increase in demand of personalized medicine. According to a report by The Insight Partners, the global genome editing market is predicted to reach approximately $ 10,691 million by 2025, increasing from $3201 million in 2017. Few factors that are fuelling the growth of the market are:
- Increased funding
- Dominance of genetic disorders
- Advancement in genome editing technology
Increased funding
The genome editing market is predicted to expand in the coming years as there is an inflow of funding from the governments in various countries. The increase in grants and funds are contributing actively to enhance genome editing research. Governments of different countries are supporting private as well as public research institutions to boost the research activities related to genome editing. For instance, In January 2018, the US government donated $190 million for research in genome editing for next 6 years. This action by the government is towards the hope to explore and develop cancer therapies and others with the help of gene editing. Additionally, the National Institutes of Health (NIH) has reserved around $45.5 million for next 4 fiscal years for the Somatic Cell Genome Editing program.
Increase in Biotech M&A deals
Private companies are moving forward to the forefront to develop gene editing technologies. Over the last year, almost 30 biotech M&A deals have been made, increasing from 20 in the year 2018. Several high-profile gene and cell therapy acquisitions took place in both public and private sectors. This trend clearly reflects the advancements made by gene technology will keep growing incrementally in the future.
In 2020, it is highly likely that large pharma companies keep acquiring and turn their focus on cell and gene therapy companies, which are focused on treatment of cancer and other rare diseases. According to a report by Deloitte, the US FDA predicts that it will approve almost 10 to 20 cell and gene therapy products a year, by 2025.
In 2020, one of the key differentiators for gene therapy companies is predicted to be manufacturing. The capacities of contract development and manufacturing organizations (CDMOs) and Contract manufacturing organizations (CMOs) are enhancing. Several large pharma players are building their own facilities as well as acquiring smaller organizations to expand their manufacturing capabilities.
Some notable investments by pharma companies:
• Catalent Inc.’s US$1.2 billion acquisition of Paragon Bioservices Inc. in Baltimore, Maryland, a viral vector CDMO for gene therapies.
• Switzerland-based Lonza Group Ltd. doubled its production capacity for viral gene and virally modified cell therapy products with a new 300,000-square-foot facility in Pearland, Texas.
• Brammer Bio is installing clinical and commercial gene therapy manufacturing capabilities at its 66,000-square-foot facility in Cambridge, Massachusetts. It was recently acquired by Thermo Fisher for US$1.7 billion.
• Precigen is adding a 5,000-square-foot facility for gene and cell manufacturing in Maryland.
• Pfizer Inc. acquired Bamboo Therapeutics in Chapel Hill, North Carolina, along with a phase I/II gene therapy manufacturing facility.
• Bluebird Bio Inc. opened its first wholly owned manufacturing facility, a 125,000-square-foot facility in Durham, North Carolina.146 Bluebird received approval from the European Medicines Agency (EMA) to manufacture its autologous gene therapy, Zynteglo, in Europe. Its CDMO is German-based apceth Biopharma GmbH, recently acquired by Hitachi Chemical. Hitachi has plans to build a regenerative medicine business in the United States, Europe, and Japan.
• Novartis is expanding its gene and cell therapy manufacturing with a new production facility in Stein, Switzerland, and adding another 38,750 square feet by acquiring CellforCure.
• Cellectis is building an 82,000-square-foot commercial manufacturing facility in North Carolina for its allogeneic CAR-T products and a 14,000-square-foot facility in Paris, France, for its allogeneic gene-edited CAR-T cell (UCART) products.
In 2020, one of the key differentiators for gene therapy companies is predicted to be manufacturing. The capacities of contract development and manufacturing organizations (CDMOs) and Contract manufacturing organizations (CMOs) are enhancing. Several large pharma players are building their own facilities as well as acquiring smaller organizations to expand their manufacturing capabilities.
What does the competition look like?
By 2023, The global contract development and manufacturing organization (CDMO) outsourcing market is predicted to reach $36.51 billion, supporting the fact that 2/3rd of the biopharmaceutical manufacturing is outsourced. According to the Deloitte global life sciences outlook 2020, the facilities of a majority of CDMO which provide gene therapies are established in the EU. Nowadays, many life science companies are adopting a new kind of outsourcing model – a hybrid model.
Financing for gene therapies
Creating value chain around patients that are pull-based should be a focus area gene therapy companies. Although most of the companies don’t have that capability yet. The commercialization of gene therapies and drug price scrutiny is highly dependent on policymakers and public. Currently, healthcare expenditures, drug pricing and market accessibility remain the top areas of concern.
Launched in 2019, the first US-approved gene therapies are Spark Therapeutics’ Luxturna and Novartis’s Zolgensma. Spark Therapeutics’ Luxturna is a treatment for a rare inherited eye disorder whereas, Novartis’s Zolgensma is a gene therapy for spinal muscular atrophy for children less than 2 years of age. In 2020, the launch of Bluebird Bio’s Zynteglo, a beta thalassemia drug is planned in the Europe.
These as well as other therapies resolve the genetic diseases but usually carry very high cost. With prices for these therapies going as high as 6 to 7 figures in dollars, the general public is not ready to accept these prices. So, it is crucial for the gene therapy companies to innovate on drug pricing and reimbursement.
With a hope that these gene therapies are mere cures rather than just treatments, $2.1 million, the cost of Zolgensma, which provides a one-time curative therapy, is almost half of the 10-year cost of current chronic management of the disease. To solve this, Novartis is partnering with insurers to formulate 5-year agreements based on successful treatment and introducing options to pay in instalments over time .
Fortunately, the prices of drugs in the United States are rising slower than ever. Many large pharma companies such as Novartis, Pfizer and Amgen didn’t opt for mid-year price increase in 2019.
What does the future hold?
With the advancement of gene technologies, scientists and researchers have been on their toes to explore what comes next. After all, it’s about the future of living beings on earth, that will impact everyone. The need for smart regulations can’t be overlooked. The amalgamation of AI and genomics was a thing of science fiction until a few years ago, and we just hit a milestone with CRISPR-Cas9. Although still at nascent stage, this innovation is revolutionary! It is here to stay and grow. In the future, gene technology and AI might cross paths in ways we have never imagined. Truly, the future of healthcare seems to be technology driven, an idea that once seemed far-off.
