Help & Support

Sanjeevini is a comprehensive, automated suite of computational tools developed at IIT Delhi, designed to accelerate the drug discovery pipeline.

Yes, our tools are free for academic and non-commercial use. We also provide free supercomputing time to students and researchers within India.

A drug target is a specific molecule in the body (usually a protein like an enzyme or receptor) that is directly involved in a disease process. A drug is designed to bind to this specific target to alter its function and produce a therapeutic effect. This interaction is often compared to a key (the drug) fitting into a specific lock (the target) to correct a problem.

If your target protein's 3D structure is unknown, our suite provides a solution with the BhageerathH+ tool. This program computationally predicts the three-dimensional structure of a protein using only its amino acid sequence. The resulting high-quality 3D model can then be used directly for subsequent drug design steps, such as docking simulations, within the Sanjeevini workflow.

An active site is a specific, pocket-like region on an enzyme or protein where the intended biological molecules (substrates) bind and a chemical reaction takes place. In drug discovery, the active site is the primary location of interest. Drugs are typically designed to fit precisely into this site, blocking it and inhibiting the protein's activity, which in turn helps to treat a disease.

A "hit" is a compound that shows the desired biological activity in an initial screening assay; think of it as a promising but unrefined starting point. A "lead" is a more advanced compound that has evolved from a hit. It has been chemically modified and optimized by scientists to improve its potency, reduce toxicity, and give it more drug-like properties, making it a much more viable candidate for becoming a new medicine.

A "binding pose" is the specific 3D orientation and conformation (shape) a drug molecule adopts when it fits into the active site of its target protein. It describes the precise positioning of the drug that allows it to form stable interactions, such as hydrogen bonds, with the target. Predicting the most favorable, lowest-energy binding pose is a critical step in computational drug design to estimate a drug's potential effectiveness.

A "docking score" is a numerical value calculated by software to predict the binding affinity (i.e., the strength of binding) between a drug molecule and its protein target. This score is typically an estimate of the binding free energy, which represents the overall stability of the drug-target complex. Yes, for most scoring functions, a more negative (lower) score indicates a stronger and more favorable predicted interaction. A lower energy value implies a more stable system.

You need to run a Molecular Dynamics (MD) simulation because docking provides a static, "snapshot" view of the best binding pose, often treating the protein as rigid. An MD simulation adds the crucial elements of motion, flexibility, and the influence of a solvent (like water) over time.

"Druggability" is the assessment of whether a biological target, typically a protein, is likely to be effectively modulated by a drug. A highly druggable target usually possesses a distinct binding site or pocket to which a small molecule can bind with high affinity and specificity.

It's normal for AADS to find multiple pockets; it automatically ranks them based on druggability scores. You should always prioritize the highest-ranked pocket for your analysis, as it is computationally the most promising. For best results, cross-reference this prediction with published scientific literature to confirm the site's biological role.

Yes, using a known active site from a homologous protein is a valid strategy, especially if the key amino acid residues in that site are highly conserved. However, you must always validate this transferred information for your specific protein. We strongly advise using our AADS tool to predict and confirm the active site on your actual target before proceeding.

There are numerous databases because each one catalogues a different slice of the vast "chemical universe." Different libraries specialize in distinct compound types, such as approved drugs for repurposing, natural products, or large collections of commercially available molecules. Using multiple, diverse databases is a key strategy that maximizes your chances of discovering novel and effective chemical scaffolds for your target.

A metalloprotein is a protein that requires a metal ion, such as iron or zinc, for its biological function, often holding it within its active site. These metal ions have unique bonding properties that standard docking algorithms cannot accurately handle. For this reason, our suite includes the specialized ParDOCK+ tool, which is designed to correctly model these critical metal-drug interactions.

Both AMBER and GROMACS are world-class MD simulation packages, and neither is universally "better." They have different core strengths and are natively associated with different force fields, making the choice a matter of user preference and specific research goals. To offer maximum flexibility, our team provides support and detailed guides for both, allowing you to use the one you're most comfortable with.

No, a computer simulation cannot definitively prove a drug will work in a person. MD simulations are powerful predictive tools that provide strong evidence for a compound's potential and help prioritize the most promising candidates. Ultimately, a drug's true efficacy and safety can only be confirmed through rigorous experimental testing and, eventually, clinical trials.

A ligand is the scientific term for any molecule that binds specifically to a biological target, such as a protein. In the "lock and key" analogy, if the protein target is the "lock," the ligand is the "key" that is being tested or designed. The goal of drug discovery is to find a ligand that binds perfectly to the target to produce a therapeutic effect.

You'll need your protein structure in the standard Protein Data Bank (.pdb) file format, which is the most common input for our tools. For the best results, we highly recommend preparing your file by removing any non-essential water molecules, ligands, or cofactors before submission.

A force field is a parameterized, classical potential energy function used to calculate the potential energy of a molecular system. It is comprised of mathematical terms describing bonded interactions (e.g., bonds, angles, dihedrals) and non-bonded interactions (van der Waals and electrostatic forces). The quality and parameterization of the chosen force field are critical for ensuring the physical accuracy of the molecular dynamics simulation.

There is no absolute number of hits to select, as the optimal quantity depends on the project's scope and resources. A standard methodology is to first isolate the top-ranking compounds from the virtual screen, typically the top 1-2% of the entire library. From this refined subset, a chemically diverse selection of 50-100 candidates is commonly chosen based on score, binding pose, and other properties for further analysis or experimental validation.

Reverse virtual screening is a computational methodology that inverts the conventional screening workflow. Instead of screening many compounds against a single protein, one compound of interest is screened against a large database of potential protein targets. This approach, implemented in our SEARCH-ML tool, is used to identify a molecule's primary biological target.

The Root Mean Square Deviation (RMSD) is a quantitative measure of the average deviation between the atoms of two superimposed molecular structures. In trajectory analysis, it is calculated over time by comparing the coordinates of the protein or ligand in each simulation frame to a reference structure. A low and stable RMSD value that reaches a plateau is a primary indicator that the system has equilibrated and, for a ligand, suggests its binding pose is stable.

The Root Mean Square Fluctuation (RMSF) is a measure that quantifies the average displacement of individual atoms or residues from their mean position over the course of a simulation. Unlike RMSD, which tracks global deviation over time, RMSF analyzes the flexibility of specific parts of the system averaged over time. Regions exhibiting high RMSF values, such as protein loops or the N/C-termini, correspond to the most mobile and flexible segments of the molecule.

To accurately model a biological system, the simulation must replicate the physiological environment. Explicit water molecules are added to solvate the protein, which is essential for correctly representing the dielectric screening and specific hydrogen bonding that govern its behavior. Counter-ions (e.g., Na⁺, Cl⁻) are then added to neutralize the system's net charge, a critical requirement for the accurate calculation of long-range electrostatic interactions.

Minimization is a static, energy-based process that resolves unfavorable steric clashes and strain in the initial structure by adjusting atomic coordinates to find a local energy minimum. Following minimization, equilibration is a dynamic simulation phase where the system's temperature and pressure are carefully adjusted to and stabilized at their target values. This two-step procedure is critical for generating a structurally sound and thermodynamically stable state before the final "production" simulation begins.

Yes, in principle, our tools can be applied to this complex problem. Developing a single "cure" is exceptionally challenging as the common cold is caused by hundreds of distinct viruses. However, you could target an essential protein (like a protease) from a specific, prevalent virus, such as a human rhinovirus. Sanjeevini could then be used to execute the workflow of screening for and optimizing potential molecules to inhibit that single target.

Absolutely not. A positive computational result indicates that a molecule is a highly promising candidate, but it is a prediction, not a confirmation. This computational hypothesis must first be validated through extensive preclinical laboratory experiments (in vitro assays and in vivo animal studies). Only compounds demonstrating safety and efficacy in these stages can be considered for advancement to human clinical trials.

A hydrogen bond (H-bond) is a highly directional, non-covalent interaction between a hydrogen atom (the donor) and a nearby electronegative atom, such as oxygen or nitrogen (the acceptor). In drug design, the formation of specific hydrogen bonds between a ligand and its protein target is a primary determinant of binding affinity and selectivity. During MD simulations, the stability and persistence of these bonds are analyzed, as they are a key indicator of a stable and promising drug-target interaction.

The Radius of Gyration (RoG) is a measure of the overall compactness of a protein's structure, calculated as the mass-weighted root mean square distance of a collection of atoms from their common center of mass. In molecular dynamics simulations, it is used to assess the stability of the protein's tertiary structure over time. A stable RoG value indicates the protein is maintaining its folded conformation, while significant fluctuations or an increasing trend can suggest structural unfolding or instability.

An allosteric site is a regulatory binding site on a protein that is physically distinct from the primary active (orthosteric) site. The binding of a molecule to this site induces a conformational change in the protein, which in turn modulates the activity of the active site. Targeting allosteric sites is a key strategy for developing drugs with higher specificity and novel mechanisms of action.

A pharmacophore is an abstract 3D representation of the essential molecular features required for a molecule to be recognized by a specific biological target. These features include hydrogen bond donors/acceptors, aromatic rings, and hydrophobic centers, arranged in a precise geometric orientation. Pharmacophore models serve as powerful 3D search queries for discovering novel and structurally diverse compounds.

ADMET is an acronym for Absorption, Distribution, Metabolism, Excretion, and Toxicity. These properties describe the disposition of a drug within an organism, governing its bioavailability, efficacy, and potential for adverse effects. Computational prediction of ADMET properties is a critical step in the early stages of drug discovery to filter out candidates likely to fail in clinical development.

Binding free energy (ΔG) is the overall change in Gibbs free energy upon the formation of a drug-target complex, representing the thermodynamic affinity of the interaction. Unlike docking scores, it is a more rigorous quantity calculated using computationally intensive methods like MMGBSA/PBSA or alchemical free energy perturbation on MD trajectories. A more negative ΔG value corresponds to a stronger, more favorable binding affinity.

Periodic Boundary Conditions (PBC) are an algorithm applied in simulations to approximate an infinite system by using a finite simulation cell. The central simulation box is replicated throughout space to form an infinite lattice, and a particle exiting one face of the box re-enters through the opposite face. This method is essential for eliminating artificial surface effects and for the accurate simulation of bulk properties.

Drug repurposing, also known as repositioning, is a drug development strategy that identifies new therapeutic uses for existing, approved drugs. Because the safety and pharmacokinetic profiles of these drugs are already well-established, this approach can significantly reduce the time and cost of development. Computational methods like reverse virtual screening are highly effective for identifying potential new targets for known drugs.

A molecular fingerprint is a digital representation of a chemical structure, typically encoded as a binary string (a series of 0s and 1s). Each position in the string represents the presence or absence of a specific substructural feature or chemical property within the molecule. Fingerprints are fundamental to cheminformatics and are used for rapid database searching, similarity assessment, and as input for machine learning models.

Lead optimization is the iterative process in drug discovery that refines a promising "lead" compound to improve its drug-like characteristics. The primary goals are to enhance biological activity and selectivity while simultaneously improving ADMET (Absorption, Distribution, Metabolism, Excretion, Toxicity) properties. This crucial phase aims to transform a lead molecule into a preclinical candidate suitable for further testing.

The Binding Affinity Prediction of Protein-Ligand (BAPPL) server computes the binding free energy for a protein-ligand complex using a robust, all-atom empirical scoring function. It offers users flexibility by providing both a direct calculation mode for fully pre-parameterized inputs and an automated workflow that assigns the necessary force field parameters before scoring.