Cargando…

Latent space of a small genetic network: Geometry of dynamics and information

The high-dimensional character of most biological systems presents genuine challenges for modeling and prediction. Here we propose a neural network–based approach for dimensionality reduction and analysis of biological gene expression data, using, as a case study, a well-known genetic network in the...

Descripción completa

Detalles Bibliográficos
Autores principales: Seyboldt, Rabea, Lavoie, Juliette, Henry, Adrien, Vanaret, Jules, Petkova, Mariela D., Gregor, Thomas, François, Paul
Formato: Online Artículo Texto
Lenguaje:English
Publicado: National Academy of Sciences 2022
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9245618/
https://www.ncbi.nlm.nih.gov/pubmed/35737842
http://dx.doi.org/10.1073/pnas.2113651119
_version_ 1784738776405245952
author Seyboldt, Rabea
Lavoie, Juliette
Henry, Adrien
Vanaret, Jules
Petkova, Mariela D.
Gregor, Thomas
François, Paul
author_facet Seyboldt, Rabea
Lavoie, Juliette
Henry, Adrien
Vanaret, Jules
Petkova, Mariela D.
Gregor, Thomas
François, Paul
author_sort Seyboldt, Rabea
collection PubMed
description The high-dimensional character of most biological systems presents genuine challenges for modeling and prediction. Here we propose a neural network–based approach for dimensionality reduction and analysis of biological gene expression data, using, as a case study, a well-known genetic network in the early Drosophila embryo, the gap gene patterning system. We build an autoencoder compressing the dynamics of spatial gap gene expression into a two-dimensional (2D) latent map. The resulting 2D dynamics suggests an almost linear model, with a small bare set of essential interactions. Maternally defined spatial modes control gap genes positioning, without the classically assumed intricate set of repressive gap gene interactions. This, surprisingly, predicts minimal changes of neighboring gap domains when knocking out gap genes, consistent with previous observations. Latent space geometries in maternal mutants are also consistent with the existence of such spatial modes. Finally, we show how positional information is well defined and interpretable as a polar angle in latent space. Our work illustrates how optimization of small neural networks on medium-sized biological datasets is sufficiently informative to capture essential underlying mechanisms of network function.
format Online
Article
Text
id pubmed-9245618
institution National Center for Biotechnology Information
language English
publishDate 2022
publisher National Academy of Sciences
record_format MEDLINE/PubMed
spelling pubmed-92456182022-12-22 Latent space of a small genetic network: Geometry of dynamics and information Seyboldt, Rabea Lavoie, Juliette Henry, Adrien Vanaret, Jules Petkova, Mariela D. Gregor, Thomas François, Paul Proc Natl Acad Sci U S A Physical Sciences The high-dimensional character of most biological systems presents genuine challenges for modeling and prediction. Here we propose a neural network–based approach for dimensionality reduction and analysis of biological gene expression data, using, as a case study, a well-known genetic network in the early Drosophila embryo, the gap gene patterning system. We build an autoencoder compressing the dynamics of spatial gap gene expression into a two-dimensional (2D) latent map. The resulting 2D dynamics suggests an almost linear model, with a small bare set of essential interactions. Maternally defined spatial modes control gap genes positioning, without the classically assumed intricate set of repressive gap gene interactions. This, surprisingly, predicts minimal changes of neighboring gap domains when knocking out gap genes, consistent with previous observations. Latent space geometries in maternal mutants are also consistent with the existence of such spatial modes. Finally, we show how positional information is well defined and interpretable as a polar angle in latent space. Our work illustrates how optimization of small neural networks on medium-sized biological datasets is sufficiently informative to capture essential underlying mechanisms of network function. National Academy of Sciences 2022-06-22 2022-06-28 /pmc/articles/PMC9245618/ /pubmed/35737842 http://dx.doi.org/10.1073/pnas.2113651119 Text en Copyright © 2022 the Author(s). Published by PNAS. https://creativecommons.org/licenses/by-nc-nd/4.0/This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND) (https://creativecommons.org/licenses/by-nc-nd/4.0/) .
spellingShingle Physical Sciences
Seyboldt, Rabea
Lavoie, Juliette
Henry, Adrien
Vanaret, Jules
Petkova, Mariela D.
Gregor, Thomas
François, Paul
Latent space of a small genetic network: Geometry of dynamics and information
title Latent space of a small genetic network: Geometry of dynamics and information
title_full Latent space of a small genetic network: Geometry of dynamics and information
title_fullStr Latent space of a small genetic network: Geometry of dynamics and information
title_full_unstemmed Latent space of a small genetic network: Geometry of dynamics and information
title_short Latent space of a small genetic network: Geometry of dynamics and information
title_sort latent space of a small genetic network: geometry of dynamics and information
topic Physical Sciences
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9245618/
https://www.ncbi.nlm.nih.gov/pubmed/35737842
http://dx.doi.org/10.1073/pnas.2113651119
work_keys_str_mv AT seyboldtrabea latentspaceofasmallgeneticnetworkgeometryofdynamicsandinformation
AT lavoiejuliette latentspaceofasmallgeneticnetworkgeometryofdynamicsandinformation
AT henryadrien latentspaceofasmallgeneticnetworkgeometryofdynamicsandinformation
AT vanaretjules latentspaceofasmallgeneticnetworkgeometryofdynamicsandinformation
AT petkovamarielad latentspaceofasmallgeneticnetworkgeometryofdynamicsandinformation
AT gregorthomas latentspaceofasmallgeneticnetworkgeometryofdynamicsandinformation
AT francoispaul latentspaceofasmallgeneticnetworkgeometryofdynamicsandinformation