# Virtual screening of compounds for the development of thyroid hormone analogues for potential application in cardiac regeneration

## TL;DR

This paper uses virtual screening to find compounds that could activate a thyroid hormone receptor linked to heart cell regeneration.

## Contribution

A novel multi-scale computational framework for identifying TRα agonists with potential cardiac regenerative applications.

## Key findings

- A virtual screening pipeline identified eight candidate compounds from 412 million in ZINC15.
- Two compounds, including Cetraxate, showed interaction energies and reactivity patterns similar to T3.
- The framework integrates MM, QM, QM/MM, and Fukui function analysis for ligand evaluation.

## Abstract

Cardiovascular diseases remain the leading global cause of mortality, largely due to the limited regenerative capacity of adult cardiac tissue. Thyroid hormones, particularly tri-iodothyronine (T3), have been shown to stimulate cardiomyocyte proliferation through activation of the thyroid hormone receptor alpha (TRα), making this receptor a promising therapeutic target. Here we report a hierarchical and consensus, multi-level virtual screening pipeline integrating Molecular Mechanics (MM), Quantum Mechanics (QM), hybrid QM/MM calculations, and reactivity analysis based on the Fukui function to identify novel TRα agonists. Starting from 412 million compounds in the ZINC15 database, physicochemical filtering, validated pharmacophore matching, and docking guided by ROC curve optimization yielded 568 candidates, from which eight compounds were selected through chemically guided visual inspection. Binding affinity was evaluated with MD/MM-PBSA, PM7, and QM/MM (B3LYP/6-31G*/CHARMM36), and complemented by Fukui reactivity mapping to rationalize protein–ligand recognition. Two ligands, including the approved drug Cetraxate, consistently showed favorable interaction energies and reactivity patterns comparable to T3, suggesting agonistic potential. This work provides a rigorous, multi-scale computational framework and identifies two mechanistically supported TRα agonist candidates for future experimental validation in cardiac regeneration.

The online version contains supplementary material available at 10.1007/s10822-026-00787-5.