Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

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

Sort by
Same author

A high-endurance DNA origami snap-through switch for functional nanoscale control.

Science robotics·2026
Same author

Distinct developmental trajectories of externally and internally generated hippocampal sequences and assemblies.

Current biology : CB·2026
Same author

Transposable elements are driving rapid adaptation of Enterococcus faecium.

Nature·2026
Same author

Diversity-generating retroelements for programmable targeted hypermutagenesis.

Nature biotechnology·2026
Same author

Constrained evolutionary funnels shape viral immune escape.

Proceedings of the National Academy of Sciences of the United States of America·2026
Same author

Origins and breadth of pairwise epistasis in an α-helix of β-lactamase TEM-1.

Nature communications·2026

Related Experiment Video

Updated: Aug 27, 2025

In Vitro Selection of Aptamers to Differentiate Infectious from Non-Infectious Viruses
12:23

In Vitro Selection of Aptamers to Differentiate Infectious from Non-Infectious Viruses

Published on: September 7, 2022

1.7K

Generative and interpretable machine learning for aptamer design and analysis of in vitro sequence selection.

Andrea Di Gioacchino1, Jonah Procyk2, Marco Molari1,3,4

  • 1Laboratoire de Physique de l'Ecole Normale Supérieure, PSL & CNRS UMR8063, Sorbonne Université, Université de Paris, Paris, France.

Plos Computational Biology
|September 29, 2022
PubMed
Summary

Restricted Boltzmann Machines (RBMs) can predict aptamer binding fitness from SELEX data. This machine learning approach identifies key sequence features and generates novel aptamers, offering a powerful tool for therapeutic and diagnostic development.

More Related Videos

Exploring Sequence Space to Identify Binding Sites for Regulatory RNA-Binding Proteins
11:34

Exploring Sequence Space to Identify Binding Sites for Regulatory RNA-Binding Proteins

Published on: August 9, 2019

6.7K
A Method for Selecting Structure-switching Aptamers Applied to a Colorimetric Gold Nanoparticle Assay
12:31

A Method for Selecting Structure-switching Aptamers Applied to a Colorimetric Gold Nanoparticle Assay

Published on: February 28, 2015

15.3K

Related Experiment Videos

Last Updated: Aug 27, 2025

In Vitro Selection of Aptamers to Differentiate Infectious from Non-Infectious Viruses
12:23

In Vitro Selection of Aptamers to Differentiate Infectious from Non-Infectious Viruses

Published on: September 7, 2022

1.7K
Exploring Sequence Space to Identify Binding Sites for Regulatory RNA-Binding Proteins
11:34

Exploring Sequence Space to Identify Binding Sites for Regulatory RNA-Binding Proteins

Published on: August 9, 2019

6.7K
A Method for Selecting Structure-switching Aptamers Applied to a Colorimetric Gold Nanoparticle Assay
12:31

A Method for Selecting Structure-switching Aptamers Applied to a Colorimetric Gold Nanoparticle Assay

Published on: February 28, 2015

15.3K

Area of Science:

  • Biotechnology
  • Computational Biology
  • Machine Learning

Background:

  • Systematic Evolution of Ligands by Exponential Enrichment (SELEX) is a common method for discovering aptamers with specific binding properties.
  • Developing efficient computational methods to analyze SELEX data and predict aptamer performance is crucial for accelerating drug discovery and diagnostics.

Purpose of the Study:

  • To investigate the utility of Restricted Boltzmann Machines (RBMs), an unsupervised neural network, for analyzing SELEX data and predicting aptamer fitness.
  • To explore the interpretability of RBM parameters for identifying functional sequence features.
  • To leverage RBMs for generating novel aptamers with desired binding characteristics.

Main Methods:

  • Training RBMs on sequence ensembles from single rounds of SELEX experiments for thrombin aptamers.
  • Correlating RBM-assigned sequence scores with experimental enrichment ratios to estimate fitness.
  • Developing two RBM training protocols (count-based and unique sequence-based) to generate high-affinity and diverse aptamers, respectively.
  • Generating novel aptamers with potential mutations using RBMs and validating them via gel shift assays.
  • Comparing RBM performance against supervised learning methods like random forests and deep neural networks.

Main Results:

  • RBMs successfully assigned scores to sequences that correlated with their experimental fitness.
  • Trained RBMs could predict the outcomes of later selection rounds.
  • Interpretable RBM parameters identified sequence features critical for aptamer binding.
  • RBMs generated novel aptamers with validated binding properties.
  • The performance of RBMs was comparable to or better than supervised learning models.

Conclusions:

  • RBMs provide an effective unsupervised approach for analyzing SELEX data and predicting aptamer binding affinity.
  • The interpretability of RBMs offers insights into sequence-fitness relationships.
  • RBMs can be utilized as a generative model to design novel aptamers with tailored properties.
  • This approach holds promise for accelerating the development of aptamer-based diagnostics and therapeutics.