R & D

what we offer

Our Capabilities

Scientific discovery, creation, design, and development of drug molecules, chemicals, formulations, polymers, catalysts, and other materials.

Sravathi offers novel drug design solutions by using a proprietary AI based solution platform to generate novel compounds, identify and validate targets. Using a diverse range of parameters, which includes physico-chemical, ADMET, pharmaco-kinetic properties, clients can rapidly down select to a few 100 compounds that can be prioritised for chemical synthesis. This reduces uncertainty in the drug discovery process, saving time and costs during development.

Generative AI

Drug Discovery

  • New Target identification
  • Off -Target information
  • New Molecule Generation
  • Property predictions

Chemistry AI

  • Route of Synthesis
  • Impurity Prediction
  • Toxicity Prediction
  • Solubility Prediction
  • Stability and Bioequivalence predictions
  • Reaction Optimisation
  • Yield Prediction


Artificial Intelligence

Artificial Intelligence capabilities for discovery using high powered in-house computing infrastructure, primarily working with own customized Deep Learning AI models.

AI capabilities include working with state of the art AI Deep Learning models for molecule discovery, activity and property predictions, and virtual screening.

Our AI solution is a platform of predictive models, optimization frameworks and an active learning workflow. In simple terms, we use AI to predict and optimize a drug’s most important properties, such as potency, selectivity, toxicity, etc. Our platform can also be used for accelerating Drug Repurposing or repositioning.

This integrated AI-to-WorkBench workflow ensures direct in vitro validation of novel in-silico target discoveries and looped back into the AI pipeline for improved optimisation, thus increasing the success rate.

Some of the other capabilities provided by our Solution platforms are:

  • Virtual molecule screening
  • In silico Target Identification
  • In silico Target Validation
  • Optimisation of Hit / Lead Candidates
  • Toxicity predictions
  • Visualise the most similar compounds of your target substance
  • NCE generation
  • Identify similar compounds of interest that may be available for patenting or for drug development.
  • Identify compounds in the chemical space that may be biologically active against a particular target of interest


Molecular Modelling

High throughput in-house virtual screening via state-of-the-art bio-informatics modelling, in-silico experiments, and perturbation modelling methods at atomic and sub-atomic levels.

Some of the capabilities developed in-house are:

  • Molecular Modelling
  • De novo Drug designing
  • Molecular Dynamic simulations
  • Property Predictions
  • FEP
  • DFT
  • QSAR


Our current pipelines primarily consist of deep learning driven NCE (New Chemical Entity), DRP (Drug Re-Purposing), reaction optimization, formulations, and allied areas.


Route of Synthesis (Retrosynthesis)

Some of the capabilities developed in-house are:

  • API
  • Intermediates
  • Small molecules
  • Agrochemicals
  • NCE Chemicals
  • Specialty/fine chemicals


Impurity Prediction

  • API / Intermediates/Small molecules
  • Nitrosamine impurities
  • Unknown to known impurities
  • Optimization of NCE/Lead /Pre-clinical candidates
  • Impurities in formulation
  • Forced degradation studies/Stability impurities


Free Energy Perturbation Calculations

Tool development
  • Performed using well-established target systems like Lysozyme, TYK2, PTP1B, and BRD2, to name few.

Tool Validation
  • Common choice of target systems with Schrodinger (used for FEP+) and arriving at similar results from our FEP tool.
  • Validation of targets like BRD2 and RBP chosen to differentiate between a weak binder and a strong binder (micromolar and nanomolar concentrations) completed successfully.
  • FEP tool validation for a) choice of FF: Amber vs OPLS, b) Number of Lambdas, c) Timing of the turn on and off of VDW and ES interactions and d) bonded interactions.

Tool Application
  • Applied the best optimized parameters obtained from validation to identify NCEs and possible repurposed candidates.
  • Predict a binder and a non-binder accurately, subsequently supported by experimental evidence.


DFT for Drug Discovery

  • ‘Density Functional Theory (DFT)’ which is a quantum mechanical method (QMM) offers the solution to most of the complex drug related problems at atomic level.
  • DFT method is employed to calculate the binding affinity of the protein-ligand interaction and aids the drug discovery pipeline to select the ligand pose.
  • Organometallics which are the part of biological system will be studied using DFT.
  • DFT is used to study the molecular properties of an isolated drug compounds which provides the electronic insights of how the drug compounds preferably bind to their targets.
  • Electronic properties from DFT relates the drug properties and biological processes of the target compounds.
  • DFT for drug discovery serves as a tool in identifying the hit compounds at less computational time.