Ligand Efficiency
Economics of Drug Design and Discovery
Ligand efficiency (LE) is a crucial metric in cheminformatics, especially within the realm of drug discovery, where it significantly influences the design and optimisation of small-molecule drugs. This metric assesses the effectiveness of a molecule that binds to its target concerning its size or atomic count. By doing so, chemists can identify and prioritise compounds that demonstrate optimal binding without the unnecessary addition of large and bulky fragments. This approach mirrors principles of business economics—maximising output while minimising input. In this context, it translates to enhancing profits or return on investment without a substantial increase in capital expenditure, which, in molecular terms, refers to the need to add more atoms to achieve optimal binding.
What is Ligand Efficiency?
Ligand efficiency (LE) measures the binding energy of a small molecule to its target, such as a protein or enzyme, per non-hydrogen atom. This concept was introduced to tackle the challenges of drug discovery, where achieving high potency often results in increased molecular size or complexity. LE ensures that compounds maintain their efficiency throughout the optimisation process, which is crucial since larger molecules can lead to poor bioavailability or increased toxicity, raising the risk of failure during clinical trials. The significance of LE is highlighted by its potential to reduce attrition rates in drug development, a major concern given the estimated $1.8 billion cost associated with bringing a new drug to market. A systematic review published in Nature Reviews Drug Discovery in 2014 noted that most marketed oral drugs tend to exhibit highly optimised LE values, underscoring their importance in successful drug development.
Why Ligand Efficiency Matters?
Optimising Drug Candidates: LE guides the selection of fragments, hits, and leads, ensuring they have the potential to become viable drugs with minimal toxicity and off-target effects.
Reducing Molecular Size/Bulkiness: By focusing on efficiency, LE helps prevent the unnecessary increase in molecular weight or lipophilicity during optimisation.
Assessing Target Druggability: LE can indicate how feasible it is to develop a drug for a specific target, aiding strategic decisions in drug discovery.
Calculating Ligand Efficiency
Ligand efficiency is calculated by normalising the binding affinity of a ligand by its size, typically measured as the number of heavy (non-hydrogen) atoms. The following equation is generally used for its calculation:
Where:
ΔG is the Gibbs free energy of binding, derived from the dissociation constant using
\(\Delta G = -RT \ln K_i\)R is the gas constant and
T is the temperature in Kelvin.
Ki is the inhibition constant.
N is the number of heavy atoms in the ligand.
For practical purposes,
LE can be approximated using the half-maximal inhibitory concentration (IC50):
This formula allows chemists to quickly assess a compound’s efficiency based on experimental data. This is the most handy way to calculate LE because Ki demands lengthy experiments for its determination compared to IC50.
Other versions of LE
To provide a more comprehensive evaluation, several related metrics are used alongside LE:
Group Efficiency (GE)
\(GE = -\frac{\Delta \Delta G}{\Delta N}\)Measures the efficiency of structural groups added during optimisation, where
ΔΔG is the change in Gibbs free energy, and
ΔN is the change in heavy atoms
Lipophilic Ligand Efficiency (LipE/LLE)
\(LipE = p\text{IC}_{50} - \log P\)Balances potency with lipophilicity, where log P is the octanol-water partition coefficient.
Binding Efficiency Index (BEI)
\(BEI = \frac{pK_i}{\text{MW}}\)Normalises potency by molecular weight in kilodaltons.
Surface-Binding Efficiency Index (SEI)
\(SEI = \frac{pK_i}{\text{PSA}\text{/100}}\)Incorporates polar surface area (PSA) to assess surface-binding efficiency. This version can be particularly important while optimising molecules that need to cross the blood-brain-barrier.
Percentage Efficiency Index (PEI)
\(PEI = \frac{\%{inhibition}}{{MW}}\)Measures inhibition relative to molecular weight, though less comparable across concentrations.
Applications:
Fragment-Based Drug Discovery (FBDD)
FBDD starts with small molecules (fragments) that bind weakly but efficiently to a target. LE is critical here because fragments must have high efficiency to be viable starting points for optimisation. A study in the Journal of Cheminformatics (2019) noted that fragments often exhibit high LE due to their efficient atom utilisation, making them ideal for developing into clinical candidates.
Hit-to-Lead Optimization
After identifying hits through high-throughput screening (HTS), LE helps prioritise compounds for further development. Metrics like LipE ensure that hits are not only potent but also have favourable properties such as low lipophilicity, which is crucial for drug-like behaviour.
Lead Optimization
During lead optimisation, LE guides the addition of functional groups to improve potency without compromising efficiency. A retrospective analysis of 46 marketed oral drugs showed that they often have highly optimised LE and LipE values, demonstrating the metric’s importance in producing successful drugs.
Virtual Screening
In virtual screening, LE is used to filter large compound libraries, prioritising those with high predicted efficiency for experimental validation. Molecular docking tools like GOLD and Glide incorporate LE scores to rank binding poses.
Target Druggability Assessment
LE can estimate a target’s druggability by analysing how efficiently ligands bind to it. A study using 480 target-assay pairs found that mean LE, combined with drug-like properties, can predict whether a target is suitable for drug development.
Tools and Software for Ligand Efficiency Analysis
Chemoinformatics relies on a variety of computational tools to calculate and analyse ligand efficiency:
Molecular Docking Software:
GOLD and Glide: These tools perform docking simulations and calculate LE scores to evaluate binding efficiency.
\(LIE = \frac{\text{GlideScore}} {\text{number of heavy atoms}} \)
Cheminformatics Platforms:
ChEMBL: A database of bioactive molecules, used to analyse LE across large datasets.
RDKit: An open-source library for calculating physicochemical properties and efficiency metrics.
Fragment-Based Design Tools:
Fragalysis: A platform for FBDD that incorporates LE to guide hit selection and optimisation.
Machine Learning and AI:
Deep learning models predict LE and related properties, enhancing virtual screening and lead optimisation.
Challenges and Future Directions
Despite its utility, ligand efficiency has limitations:
Context-Dependence: LE’s effectiveness varies by target and chemotype, requiring careful interpretation.
Metric Limitations: The Journal of Cheminformatics (2019) argues that LE is not a true metric due to its dependence on concentration units, proposing alternatives like molecular size efficiency (MSE).
Overemphasis Risk: Focusing solely on LE may neglect other critical properties like selectivity or solubility.
Future Directions
AI Integration: Combining LE with AI-driven predictions can improve the accuracy of virtual screening and lead optimisation.
Target-Specific Metrics: Developing LE metrics tailored to specific targets could enhance their applicability.
Beyond Rule-of-5: As drug discovery explores larger molecules, new efficiency metrics may be needed.