Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

Evolutionary Relationships through Genome Comparisons

Evolutionary Relationships through Genome Comparisons

Genome comparison is one of the excellent ways to interpret the evolutionary relationships between organisms. The basic principle of genome comparison is that if two species share a common feature, it is likely encoded by the DNA sequence conserved between both species. The advent of genome sequencing technologies in the late 20th century enabled scientists to understand the concept of conservation of domains between species and helped them to deduce evolutionary relationships across diverse...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Hybrid Computational Framework Integrating Ensemble Learning, Molecular Docking, and Dynamics for Predicting Antimalarial Efficacy of Malaria Box Compounds.

International journal of molecular sciences·2026

Same author

Leveraging Different Distance Functions to Predict Antiviral Peptides with Geometric Deep Learning from ESMFold-Predicted Tertiary Structures.

Antibiotics (Basel, Switzerland)·2026

Same author

StarPepWeb: an integrative, graph-based resource for bioactive peptides.

Bioinformatics advances·2025

Same author

Optimal Descriptor Subset Search via Chemical Information and Target Activity-Guided Algorithm for Antimicrobial Peptide Prediction.

Journal of chemical information and modeling·2025

Same author

Predicting Antimicrobial Peptides Using ESMFold-Predicted Structures and ESM-2-Based Amino Acid Features with Graph Deep Learning.

Journal of chemical information and modeling·2024

Same author

StarPep Toolbox: an open-source software to assist chemical space analysis of bioactive peptides and their functions using complex networks.

Bioinformatics (Oxford, England)·2023

Search research articles

Related Experiment Video

Updated: Jun 30, 2025

Creating and Applying a Reference to Facilitate the Discussion and Classification of Proteins in a Diverse Group

Creating and Applying a Reference to Facilitate the Discussion and Classification of Proteins in a Diverse Group

Published on: August 16, 2017

Examining evolutionary scale modeling-derived different-dimensional embeddings in the antimicrobial peptide

Karla L Martínez-Mauricio¹, César R García-Jacas², Greneter Cordoves-Delgado¹

¹Departamento de Ciencias de la Computación, Centro de Investigación Científica y de Educación Superior de Ensenada (CICESE), Ensenada, Mexico.

Protein Science : a Publication of the Protein Society

|March 19, 2024

Summary

This study shows that specific embeddings from Evolutionary Scale Modeling (ESM-2) models effectively classify antimicrobial peptides (AMPs). Non-deep learning Quantitative Structure-Activity Relationship (QSAR) models using these features perform comparably or better than deep learning approaches.

Keywords:

ESM-2 KNIME QSAR antimicrobial peptides deep learning ensemble classifiers evolutionary scale modeling shallow classifiers

More Related Videos

A Clinical Metaproteomics Workflow Implemented within Galaxy Bioinformatics Platform to Analyze Host-Microbiome Interactions Underlying Human Disease

A Clinical Metaproteomics Workflow Implemented within Galaxy Bioinformatics Platform to Analyze Host-Microbiome Interactions Underlying Human Disease

Published on: January 10, 2025

Antimicrobial Peptides Produced by Selective Pressure Incorporation of Non-canonical Amino Acids

Antimicrobial Peptides Produced by Selective Pressure Incorporation of Non-canonical Amino Acids

Published on: May 4, 2018

Related Experiment Videos

Last Updated: Jun 30, 2025

Creating and Applying a Reference to Facilitate the Discussion and Classification of Proteins in a Diverse Group

Creating and Applying a Reference to Facilitate the Discussion and Classification of Proteins in a Diverse Group

Published on: August 16, 2017

A Clinical Metaproteomics Workflow Implemented within Galaxy Bioinformatics Platform to Analyze Host-Microbiome Interactions Underlying Human Disease

A Clinical Metaproteomics Workflow Implemented within Galaxy Bioinformatics Platform to Analyze Host-Microbiome Interactions Underlying Human Disease

Published on: January 10, 2025

Antimicrobial Peptides Produced by Selective Pressure Incorporation of Non-canonical Amino Acids

Antimicrobial Peptides Produced by Selective Pressure Incorporation of Non-canonical Amino Acids

Published on: May 4, 2018

Area of Science:

Bioinformatics
Cheminformatics
Computational Biology
Machine Learning in Drug Discovery

Background:

Molecular features are crucial for Quantitative Structure-Activity Relationship (QSAR) modeling.
Pre-trained models like Evolutionary Scale Modeling (ESM-2) offer powerful embeddings for downstream tasks.
ESM-2 models have shown promise in protein structure prediction.

Purpose of the Study:

To evaluate the utility of ESM-2 model embeddings for classifying antimicrobial peptides (AMPs).
To compare the performance of QSAR models built using ESM-2 embeddings against state-of-the-art deep learning models.
To develop a reproducible workflow for fair comparison of computational methods.

Main Methods:

Utilized a KNIME workflow to ensure consistent methodology for QSAR model development.
Extracted embeddings from various ESM-2 models (30- and 33-layer) at different dimensions.
Built and compared QSAR models using single ESM-2 model embeddings, fused embeddings, and compared against deep learning models.

Main Results:

640-dimensional embeddings from 30-layer ESM-2 and 1280-dimensional embeddings from 33-layer ESM-2 models yielded the best QSAR model performances.
Fusing features from multiple ESM-2 models improved QSAR model performance compared to using single models.
Frequency analysis indicated that only a subset (43-66%) of ESM-2 embeddings were actively used in modeling.
Non-deep learning QSAR models, when developed with principled methodology, achieved comparable or superior performance to deep learning models for AMP prediction.

Conclusions:

Specific ESM-2 embeddings are highly valuable for QSAR-based AMP classification.
Feature fusion from multiple ESM-2 models enhances predictive power.
The developed KNIME workflow facilitates fair comparisons and the proposal of novel non-deep learning QSAR models.
Methodologically sound non-deep learning QSAR models remain competitive with deep learning approaches for AMP prediction.