We use cookies to understand how you use our site and to improve the overall user experience. This includes personalizing content and advertising. Read our
Privacy Policy
Profacgen offers computer aided drug design (CADD) services, integrating structure-based and ligand-based methodologies to accelerate lead compound identification and optimization, reducing screening costs and shortening development timelines through rational, computationally guided drug discovery.
Drug design has evolved alongside pharmaceutical chemistry. Since the 1970s, drug design has made comprehensive use of pharmaceutical chemistry, molecular biology, quantum chemistry, statistical theory, modern science and technology, and electronic computers, opening a new era in rational drug development. Computer Aided Drug Design (CADD) is a method based on computational chemistry that designs and optimizes lead compounds through computer simulation, calculation, and estimation of the relationship between a drug and its receptor biomacromolecule.
Profacgen's CADD team has successfully supported several pharmaceutical chemistry integration projects and individual projects. We bring profound scientific literacy and years of experience in the pharmaceutical industry, delivering refined expertise and practical insights to advance our clients' drug discovery programs from target validation to lead optimization.
The Role of CADD in Modern Drug Discovery
Our CADD platform delivers strategic advantages across the drug discovery pipeline:
Reduced screening costs: Virtual screening of millions of compounds against a target protein is performed computationally at a fraction of the cost of high-throughput experimental screening. CADD prioritizes the most promising candidates for synthesis and validation, minimizing wet-lab expenditure on low-probability hits
Faster lead identification: Structure-based docking and ligand-based pharmacophore screening rapidly identify hit compounds from large chemical libraries. Computational filtering by drug-likeness, ADMET properties, and synthetic accessibility further narrows the candidate pool, accelerating the transition from target to validated lead
Rational drug design: CADD provides atomic-level understanding of drug-target interactions, enabling structure-guided modification of lead compounds to improve potency, selectivity, and physicochemical properties. Rational design reduces reliance on serendipity and supports patentable chemical matter generation
Our CADD Services
Profacgen offers comprehensive CADD services spanning structure-based and ligand-based drug design paradigms:
Virtual Screening
High-throughput computational screening of chemical libraries to identify potential binders.
Structure-based virtual screening: docking of millions of compounds into target binding sites using DOCK, AutoDock, Glide, GOLD, and FlexX
Ligand-based virtual screening: pharmacophore matching, shape similarity, and fingerprint-based searching
Consensus scoring and machine learning ranking to prioritize hits for experimental validation
Molecular Docking
Prediction of binding modes and affinities for small molecules within target protein cavities.
Traditional docking: rigid and flexible receptor docking for pose prediction and affinity estimation
Reversed docking (RT-Dock): target fishing to identify potential off-targets and polypharmacology profiles
Covalent docking and metadocking for complex binding scenarios
Hit Identification
Discovery of novel chemical starting points through diverse computational strategies.
Fragment-based design: fragment searching and scaffold hopping using Ludi, MFS, and MCSS methods
Pharmacophore-based screening: DISCO and Catalyst-based feature mapping
Similarity search and clustering for lead-like compound identification
Lead Optimization
Structure-guided refinement of hit compounds to improve potency, selectivity, and developability.
Quantitative structure-activity relationship (QSAR) modeling and matched pair analysis
Potency cliff analysis and R-group decomposition for SAR elucidation
CoMFA, CoMSIA, SOMFA, and MFA for 3D-QSAR-guided optimization
Binding Mode Analysis
Detailed characterization of ligand-target interactions to guide rational modification.
Interaction fingerprinting: hydrogen bonds, hydrophobic contacts, π-π stacking, and salt bridges
Binding free energy decomposition by residue and energy component
Conformational analysis and molecular dynamics refinement of docking poses
CADD Methodologies: SBDD vs. LBDD
CADD is divided into two complementary approaches. The following comparison guides method selection:
Core structure, pharmacophore, or field-based alignment
Knowledge-Based
Target-like, active site-like, protein diversity universe
Drug-like, lead-like, diversity
Hypothetical Model
Initial design based on target structure
Hypothetical receptor model (HASL)
Figure 1. Traditional workflow of structure-based drug design (SBDD) and ligand-based drug design (LBDD). (Lu et al., 2018)
Computational Workflow
Our CADD projects follow a structured, iterative workflow from target to optimized lead:
strong>Target Structure Preparation: Target protein structures are retrieved from the PDB or generated by homology modeling. Missing loops are reconstructed, protonation states are assigned, and binding pockets are characterized by volume, shape, and physicochemical properties. Grid-based mapping identifies favorable interaction regions for ligand design.
Virtual Screening: Structure-based docking or ligand-based pharmacophore screening is performed against curated compound libraries (millions to billions of molecules). Filtering by drug-likeness (Lipinski rules, lead-likeness), ADMET properties, and synthetic accessibility reduces the candidate pool to experimentally tractable numbers.
Hit Selection and Validation: Top-ranked compounds are selected based on docking scores, interaction complementarity, and diversity. Consensus scoring and machine learning models further refine hit selection. Selected compounds are subjected to visual inspection and molecular dynamics validation to confirm binding mode stability.
Docking Analysis and Binding Mode Characterization: Detailed interaction analysis maps hydrogen bonds, hydrophobic contacts, π-π interactions, and electrostatic complementarity. Binding free energy calculations estimate affinity and guide compound ranking. Off-target profiling by reversed docking assesses selectivity liabilities.
Lead Optimization: Structure-guided modification of hit compounds improves potency, selectivity, and ADMET properties. QSAR modeling, matched pair analysis, and potency cliff analysis guide SAR development. Iterative design-synthesis-test cycles are supported by computational prioritization of candidate modifications.
Applications
Our CADD services support diverse therapeutic modalities and target classes:
Small Molecule Drug Discovery: Structure-based virtual screening, docking, and lead optimization for kinases, proteases, GPCRs, ion channels, and nuclear receptors. We support hit-to-lead and lead-to-candidate progression with rational design strategies that improve potency, selectivity, and oral bioavailability
Protein Degrader Discovery: Computational design of PROTAC linkers, E3 ligase recruiter optimization, and ternary complex modeling. Molecular dynamics simulations guide linker length and composition to maximize target engagement and ubiquitination efficiency
Enzyme Inhibitor Development: Active site mapping, transition state modeling, and mechanism-based inhibitor design. We characterize substrate binding modes, identify allosteric pockets, and predict resistance mutations to guide inhibitor development with durable efficacy
Antibody Engineering: CDR loop modeling, epitope mapping, and humanization support. We predict developability liabilities, optimize Fc effector function, and guide affinity maturation through structure-guided mutagenesis
Why Choose Profacgen
High Success Rate: Our CADD approach delivers high hit rates and success rates with relatively low research and development costs and short development cycles, maximizing return on investment for drug discovery programs.
Industry-Standard Software: We support multiple chemical software platforms including PyMOL, Schrödinger Drug Discovery Suite, CCG MOE, Dotmatics Vortex, and ChemAxon JChem Suite, ensuring access to the best tools for each task.
Experienced Team: Our CADD scientists bring profound scientific literacy and years of pharmaceutical industry experience, delivering refined expertise and practical insights to advance your programs.
Integrated Platform: We combine structure-based and ligand-based methods with molecular dynamics, QSAR, and ADMET prediction in a unified workflow, eliminating vendor coordination overhead.
Comprehensive Consultation: We provide computational chemistry and bioinformatics-related professional consultation, bridging computational predictions and experimental validation to accelerate your discovery timeline.
Scenario 1: Structure-Based Virtual Screening for Kinase Inhibitor Discovery
Program Context:
An oncology program required identification of novel inhibitors targeting a kinase with limited chemical starting points. Traditional high-throughput screening had yielded few tractable hits, and the team sought a computationally guided approach to expand the chemical diversity of the hit pool.
Objective:
To perform structure-based virtual screening of a 5-million compound library against the kinase ATP-binding site, prioritize hits by predicted affinity and selectivity, and validate top candidates through experimental binding assays.
Approach:
Profacgen prepared the kinase structure from PDB coordinates, defined the ATP-binding pocket, and performed grid-based mapping with Glide. The compound library was docked and ranked by GlideScore, followed by visual inspection of top 1,000 compounds for interaction quality. A subset of 50 compounds was selected based on diversity, drug-likeness, and synthetic accessibility. Molecular dynamics simulations were performed on the top 10 compounds to validate binding mode stability.
Outcome:
Experimental validation of 50 compounds yielded 8 confirmed binders with IC50 < 10 µM, representing a 16% hit rate. The most potent compound achieved IC50 of 180 nM with >50-fold selectivity over the closest homolog. Structure-activity relationships guided by docking analysis enabled rapid optimization to sub-nanomolar potency within 3 months.
Scenario 2: Ligand-Based Pharmacophore Design for GPCR Antagonist Optimization
Program Context:
A drug discovery program targeting a GPCR required optimization of a lead compound series with moderate potency but poor selectivity against related receptors. No high-resolution crystal structure was available for the target, precluding structure-based approaches.
Objective:
To develop a ligand-based pharmacophore model from known active compounds, guide lead optimization through 3D-QSAR analysis, and improve selectivity while maintaining potency.
Approach:
Profacgen constructed a pharmacophore model using Catalyst based on 25 known GPCR actives with diverse scaffolds. The model identified three hydrophobic features, one hydrogen bond acceptor, and one aromatic ring as essential for activity. A training set of 80 compounds was used to develop a CoMFA model with q2 = 0.72 and r2 = 0.95. Predictive mapping guided synthesis of 12 analogs targeting underexplored regions of the pharmacophore.
Outcome:
Three analogs exceeded the potency of the original lead by >10-fold, with the best compound achieving Ki = 12 nM. Selectivity against the closest homolog improved from 5-fold to >100-fold. The pharmacophore model successfully predicted activity of 8 external test compounds with <2-fold error, validating the model's predictive power.
Q: What is the difference between structure-based and ligand-based drug design?
A: Structure-based drug design (SBDD) requires a 3D structure of the target protein and uses docking, molecular dynamics, and binding site analysis to design compounds that complement the target's active site. Ligand-based drug design (LBDD) relies on known active compounds and uses pharmacophore modeling, QSAR, and similarity searching to identify new actives when the target structure is unknown. The two approaches are complementary: SBDD excels when structural information is available; LBDD is essential for targets lacking crystal structures.
Q: What software platforms does Profacgen use for CADD?
A: We support multiple industry-standard software platforms including PyMOL for visualization, Schrödinger Drug Discovery Suite for docking and virtual screening, CCG MOE for pharmacophore modeling and QSAR, Dotmatics Vortex for data analysis, and ChemAxon JChem Suite for cheminformatics. We select the optimal tool for each task based on project requirements and validation benchmarks.
Q: How accurate are virtual screening predictions?
A: Virtual screening hit rates typically range from 1-30% depending on target druggability, library quality, and method selection. Structure-based approaches with well-characterized binding sites generally achieve 10-20% hit rates. Ligand-based methods perform best when multiple diverse actives are available for model training. We improve accuracy through consensus scoring, machine learning ranking, and molecular dynamics validation of top hits. Hit rate is not the only metric; enrichment factor—the concentration of actives in the selected subset versus the full library—is equally important.
Q: Can CADD predict ADMET properties?
A: Yes. We integrate physicochemical property prediction and DMPK (drug metabolism and pharmacokinetics) modeling into our CADD workflow. This includes prediction of solubility, permeability, metabolic stability, CYP inhibition, hERG liability, and plasma protein binding. These predictions guide lead optimization to improve oral bioavailability and reduce toxicity risks early in the discovery process.
Q: What is the typical timeline for a CADD project?
A: Virtual screening campaigns typically complete within 2-4 weeks from target preparation to hit list delivery. Lead optimization projects require 1-3 months depending on iteration cycles. Structure-based design with molecular dynamics refinement may extend to 4-6 weeks. We provide customized timelines based on project scope, resource availability, and urgency. Our approach delivers high success rates with relatively short development cycles.
Q: Do I need a crystal structure to use structure-based CADD?
A: A crystal structure is ideal but not strictly required. We can generate high-quality structural models through homology modeling when the target shares >30% sequence identity with a template of known structure. For targets with no suitable templates, we can employ ligand-based methods or integrate cryo-EM density maps. Our protein structure modeling service provides reliable starting models for subsequent CADD workflows.
References:
Lu W, Zhang R, Jiang H, Zhang H, Luo C. Computer-aided drug design in epigenetics. Front Chem. 2018;6:57. doi:10.3389/fchem.2018.00057
Online Inquiry
Fill out this form and one of our experts will respond to you within one business day.
Figure 1. Traditional workflow of structure-based drug design (SBDD) and ligand-based drug design (LBDD). (Lu et al., 2018)