Skip to main content

Heliconius Butterflies Mimicry

Project Members

Tanya Berger-Wolf, David Carlyn, Harry Chao, Arpita Chowdhury, Mohannad Elhamod, Anuj Karpatne, Maksim Kholiavchenko, Mridul Khurana, Jihyung Kil, Krzysztof Kozak,Christopher Lawrence, Harish Babu Manogran, Kaiya Provost, Dan Rubenstein, Chuck Stewart, Yu Su

Project Goals

To use machine learning techniques coupled with biological information to extract traits from images of butterflies, incorporate structured and phylogenetic information similar to BGNN, and train an algorithm to reproduce expert-generated classifications. 

Project Overview

We aim to use Heliconius butterflies as they are well known for their Mullerian mimicry. Mimicry as well as the wealth of taxonomic, geographical, and genomic data that come along with Heliconius will allow us to ask complex biological questions and ultimately add to the current genetic underpinnings of Mullerian mimicry.

Readily available datasets will be used to create phylogeny and trait guided algorithms.  These data will be used to further test our algorithms, and identify which parts of the phenotype break down when subspecies hybridize. In this way we hope to identify parts of the phenotype (traits) observable by selection. These traits will then be altered to produce phenotypes that we will then take to the field to measure attack rate by predators. We hope to refine additional datasets that have low pass genomic data to go along with the image to perform genotype-phenotype analysis using machine learning.

Related Publications

Reshma R Babu, Yael Stochel, Christopher Lawrence, Daniel Rubenstein, Chuck Stewart, Wei-Lun Chao, David Carlyn, Jihyung Kil, Yu Su, Luke Song, Anuj Karpatne, Mohannad Elhamod, Krzysztof Kozak, Owen McMillan, Tanya Berger-Wolf (June 20, 2022). Understanding Mimicry in Butterflies from Images using Machine Learning. 2nd CV4Animals: Computer Vision for Animal Behavior Tracking and Modeling Workshop, in conjunction with the IEEE / CVF Computer Vision and Pattern Recognition Conference 2022

Reshma R Babu; Christopher Lawrence; Daniel Rubenstein; Chuck Stewart; Wei-Lun Chao; David Carlyn; Yu Su; Anuj Karpatne; Mohannad Elhamod; Yael Stochel; Krzysztof Kozak; Owen McMillan; Tanya Berger-Wolf; Michelle Ramirez. (2023). Imageomics Approach to Understanding Mimicry in Butterflies from Images using Machine Learning. Midwest Machine Learning Symposium. Chicago, IL.

Reshma Ramesh Babu (2023). Imageomics Approach to Understanding Visual Biological Trait Similarities using Butterfly Mimicry as a Model System.