A Cross Modal Knowledge Distillation & Data Augmentation Recipe for Improving Transcriptomics Representations through Morphological Features

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bendidi, Ihab, Mesbahi, Yassir El, Denton, Alisandra K., Suri, Karush, Kenyon-Dean, Kian, Genovesio, Auguste, Noutahi, Emmanuel
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916761948913664
author Bendidi, Ihab
Mesbahi, Yassir El
Denton, Alisandra K.
Suri, Karush
Kenyon-Dean, Kian
Genovesio, Auguste
Noutahi, Emmanuel
author_facet Bendidi, Ihab
Mesbahi, Yassir El
Denton, Alisandra K.
Suri, Karush
Kenyon-Dean, Kian
Genovesio, Auguste
Noutahi, Emmanuel
contents Understanding cellular responses to stimuli is crucial for biological discovery and drug development. Transcriptomics provides interpretable, gene-level insights, while microscopy imaging offers rich predictive features but is harder to interpret. Weakly paired datasets, where samples share biological states, enable multimodal learning but are scarce, limiting their utility for training and multimodal inference. We propose a framework to enhance transcriptomics by distilling knowledge from microscopy images. Using weakly paired data, our method aligns and binds modalities, enriching gene expression representations with morphological information. To address data scarcity, we introduce (1) Semi-Clipped, an adaptation of CLIP for cross-modal distillation using pretrained foundation models, achieving state-of-the-art results, and (2) PEA (Perturbation Embedding Augmentation), a novel augmentation technique that enhances transcriptomics data while preserving inherent biological information. These strategies improve the predictive power and retain the interpretability of transcriptomics, enabling rich unimodal representations for complex biological tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2505_21317
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Cross Modal Knowledge Distillation & Data Augmentation Recipe for Improving Transcriptomics Representations through Morphological Features
Bendidi, Ihab
Mesbahi, Yassir El
Denton, Alisandra K.
Suri, Karush
Kenyon-Dean, Kian
Genovesio, Auguste
Noutahi, Emmanuel
Machine Learning
Artificial Intelligence
Understanding cellular responses to stimuli is crucial for biological discovery and drug development. Transcriptomics provides interpretable, gene-level insights, while microscopy imaging offers rich predictive features but is harder to interpret. Weakly paired datasets, where samples share biological states, enable multimodal learning but are scarce, limiting their utility for training and multimodal inference. We propose a framework to enhance transcriptomics by distilling knowledge from microscopy images. Using weakly paired data, our method aligns and binds modalities, enriching gene expression representations with morphological information. To address data scarcity, we introduce (1) Semi-Clipped, an adaptation of CLIP for cross-modal distillation using pretrained foundation models, achieving state-of-the-art results, and (2) PEA (Perturbation Embedding Augmentation), a novel augmentation technique that enhances transcriptomics data while preserving inherent biological information. These strategies improve the predictive power and retain the interpretability of transcriptomics, enabling rich unimodal representations for complex biological tasks.
title A Cross Modal Knowledge Distillation & Data Augmentation Recipe for Improving Transcriptomics Representations through Morphological Features
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2505.21317