Aganitha’s EAGLE™ Pipeline Accelerates High-Throughput In Silico Formulations

Solving Cyclic Peptide Solubility

Cyclic peptides are rapidly emerging as a transformative therapeutic modality, offering the unique ability to target challenging protein interfaces that remain inaccessible to traditional small molecules and biologics. However, despite their growing promise across oncology, immunology, infectious diseases, and metabolic disorders, poor and highly variable solubility remains a major barrier to successful development. Solubility directly influences formulation feasibility, bioavailability, manufacturability, and clinical dosing, yet experimental assessment across multiple solvent systems is costly, time-consuming, and difficult to scale for the large libraries generated in modern peptide discovery programs.

Compounding this challenge, cyclic peptide solubility is intrinsically linked to conformational dynamics. Unlike conventional drug molecules, cyclic peptides exist as ensembles of interconverting conformations whose interactions with different solvent environments govern solubility, aggregation propensity, and developability. Aganitha’s EAGLE™ pipeline addresses this critical gap by combining advanced computational modeling and high-throughput in silico formulation screening to predict solvent-dependent behavior at scale. By enabling rapid evaluation of cyclic peptide solubility across diverse formulation conditions, EAGLE™ helps accelerate candidate selection, reduce experimental burden, and de-risk downstream development.

Overview

Accurately predicting the thermodynamic properties of macrocyclic therapeutic peptides like Cyclosporin A is a critical bottleneck in computational drug formulation. Traditional quantitative structure-activity relationship (QSAR) models often fail when applied to complex cyclic peptides because they cannot capture the molecule’s hyper-flexible three-dimensional shape. To overcome this, an advanced in silico screening framework was deployed to calculate and benchmark the solubility of Cyclosporin A across a diverse range of organic and aqueous solvents, successfully matching simulated thermodynamic behavior against established experimental data.

Design Challenges

• – Conformational Complexity: Macrocycles possess highly flexible backbones with massive low-energy conformational spaces, making accurate structural initialization difficult.
• – Algorithmic Convergence Shifts: Standard iterative thermodynamic loops frequently overestimate equilibrium values, necessitating a delicate balance between infinite dilution and saturated state equations.
• – Solid-State Parameter Sensitivity: Computational outputs are highly sensitive to initial thermodynamic constants (fusion temperature and enthalpy), where minor baseline input deviations skew absolute solubility calculations.

Objectives

• – Map Multi-Solvent Profiles: Evaluate and rank the thermodynamic solubility of Cyclosporin A across water, ethanol, DMSO, chloroform, and acetone.
• – Validate Against Empirical Benchmarks: Ensure the calculated relative solubility order across diverse fluid media accurately aligns with experimental trends.
• – Optimize Computational Throughput: Refine quantum mechanical geometry optimization parameters to establish a scalable, cost-effective workflow for future cyclic peptide pipelines.

Outcomes

• – Replicated relative experimental rankings, correctly identifying aqueous media as the poorest solvent.
• – Demonstrated infinite dilution models provide superior qualitative trend tracking for macrocycles.
• – Established optimized structural ensemble density parameters to minimize predictive baseline variations.
• – Delivered a high-throughput computational pipeline that significantly reduces trial bench chemistry.

Future Directions

To build upon this framework, next steps will focus on automating the direct conversion and optimization of raw SMILES strings into quantum-ready coordinates, bypassing intermediate file dependencies. Additionally, the geometric search space within the EAGLE^TM conformer generation pipeline will be expanded by adjusting structural sampling limits, while simultaneously fine-tuning quantum-mechanical basis sets and exchange-correlation functionals to narrow the accuracy gap between iterative simulations and physical benchtop data.

More on Aganitha’s EAGLE^TM Pipeline: Click here.