TCS Drug Design & Engineering
Concept video on TCS PoV on Drug Design & Engineering
Industry
Life Sciences
Patient-centered Care Models Taking Center Stage in Life Sciences
HIGHLIGHTS
- TCS has developed a de novo drug design pipeline that uses multiple proprietary AI-based approaches to design diversified new chemical entities with desired drug-like properties by exploring the vast chemical space.
- TCS is leveraging agentic AI-based architecture to drive drug discovery transformation with scientists in loop to make faster decisions resulting in higher throughput at lower costs and improve overall scientific productivity.
- TCS is the transformation partner for multiple clients enabling Drug Discovery organizations to manage their technology landscape and help them identify and prioritize Design & Engineering opportunities.
Overview
The drug discovery process followed currently is slow, expensive, and iterative.
This calls for a shift in the way the industry has been operating to speed up the process and make it cheaper. Artificial intelligence (AI) provides an opportunity to lower costs and reduce the time taken for discovering new drugs by reducing the attrition of candidates at each stage and ensuring higher success rates from the beginning.
TCS applies AI-based methods to design molecules with desired properties and to predict physicochemical properties. These methods are borne out of TCS’ long-standing presence in genomics and drug discovery that has resulted in cutting-edge capabilities for genome interpretation, precision oncology, microbiome, and AI-based molecule design, among others. The design molecules methods are generative models used to develop new molecule design for known or unknown family of protein (target).
This is critical in the development of new medications to treat, cure, or prevent various diseases, saving and impacting the lives of billions of people.
Solution
TCS has developed a proprietary generative AI (GenAI) solution, which includes ligand-based, structure-based, and gene-expression-based approaches for novel synthesizable molecule design.
Ligand-based generative model
TCS’ ligand-based generative model grasps the rules guiding the design of small molecules. The method employs autoencoder pre-trained on millions of molecules and fine-tuned using known actives.
Structure-based generative model
TCS has developed a structure-based generative model that takes the 3D protein pocket as input and generates molecules predicted to bind with high affinity. Using a geometric encoder for pocket representation and a VAE-based decoder for molecule generation, our model captures spatial constraints and chemical complementarity. It achieves better binding affinity and other properties for the target protein of interest.
Gene-expression-based generative model
TCS has developed a gene-expression conditioned generative model trained on the LINCS L1000 dataset. We have taken special care to reduce the noise in gene expression data. It maps transcriptional signatures to a latent space aligned with molecular embeddings, enabling the generation of compounds likely to induce desired gene expression patterns.
Synthesis-aware generative model
TCS has developed a synthesis-aware generative model that co-optimizes for target properties and chemical synthesizability. A template-based forward synthesis route prediction and molecule generation is performed in parallel using the Monte Carlo Tree Search (MCTS) method where, the subsequent fragments for molecule growth can again be obtained from a purchasable fragment library.
Benefits
TCS is accelerating innovation and bringing life-saving treatments to patients faster than ever before.
Our offering provides multiple benefits such as:
- Greater and increasing rate of success of drug, molecule discovery
- Significant savings in cost and time, resulting in better profit margins for companies
- Faster drug-development lifecycle resulting in faster time to market
TCS advantage
TCS is driving transformation from drug discovery to drug design and engineering by leveraging data and AI at scale.
Our patented AI-based drug design methods and models provide a unique value proposition to reduce turn-around time and increase probability of success significantly in the early-stage drug discovery. TCS also has an experienced multidisciplinary team of researchers, domain experts and a rich co-innovation ecosystem.
Our differentiators include:
- Deep domain expertise in pharma, biopharma, quantum chemistry, chem- and bioinformatics for molecules design
- Strong IP portfolio with seven patents and over 10 publications
- Expertise to develop and train GenAI models for molecule design
