Embeddings from deep learning transfer GO annotations beyond homology

Abstract Knowing protein function is crucial to advance molecular and medical biology, yet experimental function annotations through the Gene Ontology (GO) exist for fewer than 0.5% of all known proteins. Computational methods bridge this sequence-annotation gap typically through homology-based anno...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Maria Littmann, Michael Heinzinger, Christian Dallago, Tobias Olenyi, Burkhard Rost
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/30047864d85f4054b8b99879adff217c
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:30047864d85f4054b8b99879adff217c
record_format	dspace
spelling	oai:doaj.org-article:30047864d85f4054b8b99879adff217c2021-12-02T14:12:45ZEmbeddings from deep learning transfer GO annotations beyond homology10.1038/s41598-020-80786-02045-2322https://doaj.org/article/30047864d85f4054b8b99879adff217c2021-01-01T00:00:00Zhttps://doi.org/10.1038/s41598-020-80786-0https://doaj.org/toc/2045-2322Abstract Knowing protein function is crucial to advance molecular and medical biology, yet experimental function annotations through the Gene Ontology (GO) exist for fewer than 0.5% of all known proteins. Computational methods bridge this sequence-annotation gap typically through homology-based annotation transfer by identifying sequence-similar proteins with known function or through prediction methods using evolutionary information. Here, we propose predicting GO terms through annotation transfer based on proximity of proteins in the SeqVec embedding rather than in sequence space. These embeddings originate from deep learned language models (LMs) for protein sequences (SeqVec) transferring the knowledge gained from predicting the next amino acid in 33 million protein sequences. Replicating the conditions of CAFA3, our method reaches an Fmax of 37 ± 2%, 50 ± 3%, and 57 ± 2% for BPO, MFO, and CCO, respectively. Numerically, this appears close to the top ten CAFA3 methods. When restricting the annotation transfer to proteins with < 20% pairwise sequence identity to the query, performance drops (Fmax BPO 33 ± 2%, MFO 43 ± 3%, CCO 53 ± 2%); this still outperforms naïve sequence-based transfer. Preliminary results from CAFA4 appear to confirm these findings. Overall, this new concept is likely to change the annotation of proteins, in particular for proteins from smaller families or proteins with intrinsically disordered regions.Maria LittmannMichael HeinzingerChristian DallagoTobias OlenyiBurkhard RostNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-14 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Maria Littmann Michael Heinzinger Christian Dallago Tobias Olenyi Burkhard Rost Embeddings from deep learning transfer GO annotations beyond homology
description	Abstract Knowing protein function is crucial to advance molecular and medical biology, yet experimental function annotations through the Gene Ontology (GO) exist for fewer than 0.5% of all known proteins. Computational methods bridge this sequence-annotation gap typically through homology-based annotation transfer by identifying sequence-similar proteins with known function or through prediction methods using evolutionary information. Here, we propose predicting GO terms through annotation transfer based on proximity of proteins in the SeqVec embedding rather than in sequence space. These embeddings originate from deep learned language models (LMs) for protein sequences (SeqVec) transferring the knowledge gained from predicting the next amino acid in 33 million protein sequences. Replicating the conditions of CAFA3, our method reaches an Fmax of 37 ± 2%, 50 ± 3%, and 57 ± 2% for BPO, MFO, and CCO, respectively. Numerically, this appears close to the top ten CAFA3 methods. When restricting the annotation transfer to proteins with < 20% pairwise sequence identity to the query, performance drops (Fmax BPO 33 ± 2%, MFO 43 ± 3%, CCO 53 ± 2%); this still outperforms naïve sequence-based transfer. Preliminary results from CAFA4 appear to confirm these findings. Overall, this new concept is likely to change the annotation of proteins, in particular for proteins from smaller families or proteins with intrinsically disordered regions.
format	article
author	Maria Littmann Michael Heinzinger Christian Dallago Tobias Olenyi Burkhard Rost
author_facet	Maria Littmann Michael Heinzinger Christian Dallago Tobias Olenyi Burkhard Rost
author_sort	Maria Littmann
title	Embeddings from deep learning transfer GO annotations beyond homology
title_short	Embeddings from deep learning transfer GO annotations beyond homology
title_full	Embeddings from deep learning transfer GO annotations beyond homology
title_fullStr	Embeddings from deep learning transfer GO annotations beyond homology
title_full_unstemmed	Embeddings from deep learning transfer GO annotations beyond homology
title_sort	embeddings from deep learning transfer go annotations beyond homology
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/30047864d85f4054b8b99879adff217c
work_keys_str_mv	AT marialittmann embeddingsfromdeeplearningtransfergoannotationsbeyondhomology AT michaelheinzinger embeddingsfromdeeplearningtransfergoannotationsbeyondhomology AT christiandallago embeddingsfromdeeplearningtransfergoannotationsbeyondhomology AT tobiasolenyi embeddingsfromdeeplearningtransfergoannotationsbeyondhomology AT burkhardrost embeddingsfromdeeplearningtransfergoannotationsbeyondhomology
_version_	1718391776988889088

Embeddings from deep learning transfer GO annotations beyond homology

Ejemplares similares