Bengaluru start-up exploring a sustainable drug discovery model for cancer treatment

Antibody therapy for cancer treatment, a more precise and less harmful alternative to chemotherapy, has been around for many years. However, for the common man, it remains prohibitively expensive at a cost of around ₹4 to 5 crores. Bengaluru-based startup ImmunitoAI may have a solution.

Founded by Dr. Aridani Shah and Trisha Chatterjee, ImmunitoAI is building an AI platform for antibody-based drug discovery that makes designing and developing antibody drugs faster and more efficient.

chemotherapy vs antibody therapy

According to the report the market size of antibody therapy is as large as $186 billion (by 2021) and is expected to witness a CAGR of 13.2 percent from 2022 to 2028. A relatively new market, Dr. Shah points out that five of the top 10 best sellers in medicine today are antibodies.

Chemotherapy drugs, which are small molecule drugs, kill any rapidly growing cell, and cannot differentiate between a healthy cell and a cancerous cell. As a result, they damage every rapidly growing cell in the body, including hair and red blood cells.

On the other hand, antibodies bind to a specific target. Using this property, pharma companies are designing antibody-based drugs that can bind to cancer cells. These antibodies will activate the immune system which will only attack the cancer cell.

“Since this is a biological molecule, until now companies have relied on biological sources such as animals to obtain these antibodies,” explains Ms. Shah.

For example, some proteins are specifically present only on cancer surfaces. You can just take that protein and inject it into an animal. For the animal, it is a foreign body. Therefore, its immune system starts producing antibodies over the course of three months to a year.”

Then the cells that make the antibodies are extracted. Millions of antibodies are screened to identify the antibodies that bind to the target. Medicines are made from these antibodies, called lead molecules.

However, there’s a catch.

trial-and-error method

“Pharma companies have put enormous effort into characterizing this molecule. Characterization means finding out what are its properties, what is the protein sequence, if it can remain stable at different temperatures, and whether it is possible to store it for a very long time…,” says Dr. Shah. This takes about three to six months.

Next is a trial-and-error phase where mutations are made to the molecule to see if it can acquire the required drug properties.

The cycle of characterization, modification, and testing continues until a satisfactory molecule is obtained.

But another challenge here is that there is a limit to changing the molecule because its source is an animal system. After a point, it is likely to lose either its function or its structure.

At this point, the pharma company has only two options – abandon the project that cost them years and millions of dollars, or go ahead with the compromise.

“If you compromise, there is a high chance that the drug will fail in animal studies or human studies,” says Dr. Shah.

AI-Designed Antibodies

The reason for such limitations is primarily the reliance on a biological source to obtain the molecule, which was first created to live inside an animal’s body, not to be a drug. But what if antibodies with the required drug properties could be designed with the help of AI and then grown in the laboratory?

Trisha Chatterjee (left) and Dr. Aridani Shah, co-founders of ImmunitoAI

With this thought process, ImmunitoAI began building an AI platform with two parts.

The first part, called imDESIGN, designs a new antibody from scratch with all the required properties. Several thousand trial-and-error procedures over the years and advances in the field have already generated a large store of sequencing data with which the startup is training its deep learning models.

“We have a generative model that generates lots of antibodies for a given antigen. Then we take it through another pipeline called imRANK,” says Dr. Shah.

“We generate a set of antibody sequences, fold it and then dock it exactly in the region where we want it to bind. Then we study all the interactions there based on which they will be ranked as good binders or weak binders.

Even the whole process is done in a computational way. Now it goes to the lab. DNA corresponding to the selected antibody design is synthesized and inserted into bacterial or mammalian cells. The cells now begin to produce the actual antibodies and those antibodies are characterized.

“Since we have already predicted the properties of the drug, characterization is no longer an exploratory path or trial and error. At this stage, we only verify whether the estimated parameters are correct or not,” says Dr. Shah.

The startup is currently training its models and has got 80 per cent accuracy in terms of getting biologically viable sequences. The next step will be to improve binding and specificity.

a sustainable model

According to Dr. Shah, once the AI ​​model is trained, it will only take a few minutes to design an antibody with the required drug properties. It will take about six to eight months to experimentally verify the molecules in the laboratory.

She notes that it will take a year or so to get the last molecule back into place. Currently, it takes the industry four to 10 years to achieve similar results.

While speed is one aspect of it, what is more important is efficiency.

She says, “Imagine a pharma company that has ten molecules in its pipeline. Each clinical trial results in a loss of approximately $500 million. Perhaps one of these molecules will eventually reach the market. The company would like to recover all its losses from the other nine. Then the cost of the medicine becomes very high.”

This has put antibody therapy out of reach of the common man.

Dr. Shah acknowledges that the ImmunoAI intervention alone may not bring down the costs drastically, but it will certainly increase the success rate significantly, at least 2-fold.

The goal is to eventually create a more sustainable model of developing the molecules, which they hope will eventually lead the way to making antibody drugs more available and accessible, similar to small molecule drugs like crocin.

the challenges

In 2021 the company raised ₹1 million in a seed round led by Pi Ventures. Existing investor Entrepreneur First was also a part of this round.

Rupan Aulakh, Managing Director, PI Ventures, said that advances in technologies such as AI are transforming drug discovery and ImmunitoAI has been a pioneer in this approach, by using AI to accelerate the pace of antibody discovery.

“This could translate into a major impact since antibodies, although very promising candidates for various therapies, are today hampered by the costly, lengthy and inefficient discovery process. The opportunity that ImmunitoAI offers is effective and safe treatment of diseases such as cancer and autoimmune diseases through innovative antibodies,” said Ms. Aulakh.

“At PI Ventures, we support deep-tech startups leveraging disruptive technologies to change the status quo or create new markets. ImmunitoAI dovetails well with this approach. They’re solving a difficult problem and building valuable IP in the process.”

The ImmunitoAI team hopes to have the final model ready in two to three months and then start generating sequences. However, the road ahead is not without challenges.

Dr. Shah says, “In biology, AI has unfortunately been misused. Many have made massive claims and not lived up to them. That’s why people are skeptical about accepting AI-based solutions in biology.”

She adds that she is also often faced with the question of FDA approval.

“Ultimately a biomolecule will be made and it will go through all the tests required by the FDA. But we’re getting to an earlier stage than that which is about identifying a good antibody. There are no rules there. Everything will come after actual animal testing,” she clarifies.

road ahead

The team has completed a set of validations and next aims to prove that the platform can generate antibody designs that bind specifically to targets.

Dr. Shah says, “We want to benchmark our antibody design against existing antibodies. We’ll take an existing target that has an existing antibody in the market. And then we will do our production. This would help to compare how many years it took for the current drug to develop and how long it took for us.

She adds, “We are first going to look at four to five cancer targets for thorough benchmarking. After that, the pharma companies we have partnered with will define the disease and the target and then the drugs will be developed.