1
|
Heinzinger M, Rost B. Artificial Intelligence Learns Protein Prediction. Cold Spring Harb Perspect Biol 2024; 16:a041458. [PMID: 38858069 PMCID: PMC11368192 DOI: 10.1101/cshperspect.a041458] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 06/12/2024]
Abstract
From AlphaGO over StableDiffusion to ChatGPT, the recent decade of exponential advances in artificial intelligence (AI) has been altering life. In parallel, advances in computational biology are beginning to decode the language of life: AlphaFold2 leaped forward in protein structure prediction, and protein language models (pLMs) replaced expertise and evolutionary information from multiple sequence alignments with information learned from reoccurring patterns in databases of billions of proteins without experimental annotations other than the amino acid sequences. None of those tools could have been developed 10 years ago; all will increase the wealth of experimental data and speed up the cycle from idea to proof. AI is affecting molecular and medical biology at giant steps, and the most important might be the leap toward more powerful protein design.
Collapse
Affiliation(s)
- Michael Heinzinger
- Technical University of Munich (TUM) School of School of Computation, Information and Technology (CIT), Bioinformatics and Computational Biology - i12, 85748 Garching/Munich, Germany
| | - Burkhard Rost
- Technical University of Munich (TUM) School of School of Computation, Information and Technology (CIT), Bioinformatics and Computational Biology - i12, 85748 Garching/Munich, Germany
- Institute for Advanced Study (TUM-IAS), 85748 Garching/Munich, Germany
- TUM School of Life Sciences Weihenstephan (WZW), 85354 Freising, Germany
- Department of Biochemistry and Molecular Biophysics, Columbia University, New York, New York 10032, USA
| |
Collapse
|
2
|
Bordin N, Scholes H, Rauer C, Roca-Martínez J, Sillitoe I, Orengo C. Clustering protein functional families at large scale with hierarchical approaches. Protein Sci 2024; 33:e5140. [PMID: 39145441 PMCID: PMC11325189 DOI: 10.1002/pro.5140] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/27/2024] [Revised: 07/22/2024] [Accepted: 07/24/2024] [Indexed: 08/16/2024]
Abstract
Proteins, fundamental to cellular activities, reveal their function and evolution through their structure and sequence. CATH functional families (FunFams) are coherent clusters of protein domain sequences in which the function is conserved across their members. The increasing volume and complexity of protein data enabled by large-scale repositories like MGnify or AlphaFold Database requires more powerful approaches that can scale to the size of these new resources. In this work, we introduce MARC and FRAN, two algorithms developed to build upon and address limitations of GeMMA/FunFHMMER, our original methods developed to classify proteins with related functions using a hierarchical approach. We also present CATH-eMMA, which uses embeddings or Foldseek distances to form relationship trees from distance matrices, reducing computational demands and handling various data types effectively. CATH-eMMA offers a highly robust and much faster tool for clustering protein functions on a large scale, providing a new tool for future studies in protein function and evolution.
Collapse
Affiliation(s)
- Nicola Bordin
- Institute of Structural and Molecular Biology, University College London, London, UK
| | - Harry Scholes
- Institute of Structural and Molecular Biology, University College London, London, UK
| | - Clemens Rauer
- Institute of Structural and Molecular Biology, University College London, London, UK
- Universidad Autonoma de Madrid, Ciudad Universitaria de Cantoblanco, Madrid, Spain
| | - Joel Roca-Martínez
- Institute of Structural and Molecular Biology, University College London, London, UK
| | - Ian Sillitoe
- Institute of Structural and Molecular Biology, University College London, London, UK
| | - Christine Orengo
- Institute of Structural and Molecular Biology, University College London, London, UK
| |
Collapse
|
3
|
Hong L, Hu Z, Sun S, Tang X, Wang J, Tan Q, Zheng L, Wang S, Xu S, King I, Gerstein M, Li Y. Fast, sensitive detection of protein homologs using deep dense retrieval. Nat Biotechnol 2024:10.1038/s41587-024-02353-6. [PMID: 39123049 DOI: 10.1038/s41587-024-02353-6] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/15/2023] [Accepted: 07/12/2024] [Indexed: 08/12/2024]
Abstract
The identification of protein homologs in large databases using conventional methods, such as protein sequence comparison, often misses remote homologs. Here, we offer an ultrafast, highly sensitive method, dense homolog retriever (DHR), for detecting homologs on the basis of a protein language model and dense retrieval techniques. Its dual-encoder architecture generates different embeddings for the same protein sequence and easily locates homologs by comparing these representations. Its alignment-free nature improves speed and the protein language model incorporates rich evolutionary and structural information within DHR embeddings. DHR achieves a >10% increase in sensitivity compared to previous methods and a >56% increase in sensitivity at the superfamily level for samples that are challenging to identify using alignment-based approaches. It is up to 22 times faster than traditional methods such as PSI-BLAST and DIAMOND and up to 28,700 times faster than HMMER. The new remote homologs exclusively found by DHR are useful for revealing connections between well-characterized proteins and improving our knowledge of protein evolution, structure and function.
Collapse
Affiliation(s)
- Liang Hong
- Department of Computer Science and Engineering, The Chinese University of Hong Kong, Hong Kong SAR, China
| | - Zhihang Hu
- Department of Computer Science and Engineering, The Chinese University of Hong Kong, Hong Kong SAR, China
| | - Siqi Sun
- Research Institute of Intelligent Complex Systems, Fudan University, Shanghai, China.
- Shanghai AI Laboratory, Shanghai, China.
| | - Xiangru Tang
- Department of Computer Science, Yale University, New Haven, CT, USA
| | - Jiuming Wang
- Department of Computer Science and Engineering, The Chinese University of Hong Kong, Hong Kong SAR, China
- OneAIM Ltd., Hong Kong SAR, China
| | - Qingxiong Tan
- Department of Computer Science and Engineering, The Chinese University of Hong Kong, Hong Kong SAR, China
| | - Liangzhen Zheng
- Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
- Shanghai Zelixir Biotech Company Ltd., Shanghai, China
| | - Sheng Wang
- Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
- Shanghai Zelixir Biotech Company Ltd., Shanghai, China
| | - Sheng Xu
- Research Institute of Intelligent Complex Systems, Fudan University, Shanghai, China
- Shanghai AI Laboratory, Shanghai, China
| | - Irwin King
- Department of Computer Science and Engineering, The Chinese University of Hong Kong, Hong Kong SAR, China
| | - Mark Gerstein
- Department of Computer Science, Yale University, New Haven, CT, USA.
- Computational Biology and Bioinformatics Program, Yale University, New Haven, CT, USA.
- Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, CT, USA.
- Department of Statistics and Data Science, Yale University, New Haven, CT, USA.
| | - Yu Li
- Department of Computer Science and Engineering, The Chinese University of Hong Kong, Hong Kong SAR, China.
- Shanghai AI Laboratory, Shanghai, China.
- Wyss Institute for Biologically Inspired Engineering, Harvard University, Boston, MA, USA.
- Institute for Medical Engineering and Science, Massachusetts Institute of Technology, Cambridge, MA, USA.
- Broad Institute of MIT and Harvard, Cambridge, MA, USA.
- The Chinese University of Hong Kong Shenzhen Research Institute, Shenzhen, China.
| |
Collapse
|
4
|
de Crécy-Lagard V, Dias R, Friedberg I, Yuan Y, Swairjo MA. Limitations of Current Machine-Learning Models in Predicting Enzymatic Functions for Uncharacterized Proteins. BIORXIV : THE PREPRINT SERVER FOR BIOLOGY 2024:2024.07.01.601547. [PMID: 39005379 PMCID: PMC11244979 DOI: 10.1101/2024.07.01.601547] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 07/16/2024]
Abstract
Thirty to seventy percent of proteins in any given genome have no assigned function and have been labeled as the protein "unknownme". This large knowledge gap prevents the biological community from fully leveraging the plethora of genomic data that is now available. Machine-learning approaches are showing some promise in propagating functional knowledge from experimentally characterized proteins to the correct set of isofunctional orthologs. However, they largely fail to predict enzymatic functions unseen in the training set, as shown by dissecting the predictions made for 450 enzymes of unknown function from the model bacteria Escherichia coli using the DeepECTransformer platform. Lessons from these failures can help the community develop machine-learning methods that assist domain experts in making testable functional predictions for more members of the uncharacterized proteome.
Collapse
|
5
|
Urhan A, Cosma BM, Earl AM, Manson AL, Abeel T. SAFPred: synteny-aware gene function prediction for bacteria using protein embeddings. Bioinformatics 2024; 40:btae328. [PMID: 38775729 PMCID: PMC11147799 DOI: 10.1093/bioinformatics/btae328] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/16/2023] [Revised: 04/08/2024] [Accepted: 05/21/2024] [Indexed: 06/04/2024] Open
Abstract
MOTIVATION Today, we know the function of only a small fraction of the protein sequences predicted from genomic data. This problem is even more salient for bacteria, which represent some of the most phylogenetically and metabolically diverse taxa on Earth. This low rate of bacterial gene annotation is compounded by the fact that most function prediction algorithms have focused on eukaryotes, and conventional annotation approaches rely on the presence of similar sequences in existing databases. However, often there are no such sequences for novel bacterial proteins. Thus, we need improved gene function prediction methods tailored for bacteria. Recently, transformer-based language models-adopted from the natural language processing field-have been used to obtain new representations of proteins, to replace amino acid sequences. These representations, referred to as protein embeddings, have shown promise for improving annotation of eukaryotes, but there have been only limited applications on bacterial genomes. RESULTS To predict gene functions in bacteria, we developed SAFPred, a novel synteny-aware gene function prediction tool based on protein embeddings from state-of-the-art protein language models. SAFpred also leverages the unique operon structure of bacteria through conserved synteny. SAFPred outperformed both conventional sequence-based annotation methods and state-of-the-art methods on multiple bacterial species, including for distant homolog detection, where the sequence similarity to the proteins in the training set was as low as 40%. Using SAFPred to identify gene functions across diverse enterococci, of which some species are major clinical threats, we identified 11 previously unrecognized putative novel toxins, with potential significance to human and animal health. AVAILABILITY AND IMPLEMENTATION https://github.com/AbeelLab/safpred.
Collapse
Affiliation(s)
- Aysun Urhan
- Delft Bioinformatics Lab, Delft University of Technology Van Mourik, Delft XE 2628, The Netherlands
- Infectious Disease and Microbiome Program, Broad Institute of MIT and Harvard, Cambridge, MA 02142, United States
| | - Bianca-Maria Cosma
- Delft Bioinformatics Lab, Delft University of Technology Van Mourik, Delft XE 2628, The Netherlands
| | - Ashlee M Earl
- Infectious Disease and Microbiome Program, Broad Institute of MIT and Harvard, Cambridge, MA 02142, United States
| | - Abigail L Manson
- Infectious Disease and Microbiome Program, Broad Institute of MIT and Harvard, Cambridge, MA 02142, United States
| | - Thomas Abeel
- Delft Bioinformatics Lab, Delft University of Technology Van Mourik, Delft XE 2628, The Netherlands
- Infectious Disease and Microbiome Program, Broad Institute of MIT and Harvard, Cambridge, MA 02142, United States
| |
Collapse
|
6
|
Hamamsy T, Morton JT, Blackwell R, Berenberg D, Carriero N, Gligorijevic V, Strauss CEM, Leman JK, Cho K, Bonneau R. Protein remote homology detection and structural alignment using deep learning. Nat Biotechnol 2024; 42:975-985. [PMID: 37679542 PMCID: PMC11180608 DOI: 10.1038/s41587-023-01917-2] [Citation(s) in RCA: 10] [Impact Index Per Article: 10.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/07/2022] [Accepted: 07/26/2023] [Indexed: 09/09/2023]
Abstract
Exploiting sequence-structure-function relationships in biotechnology requires improved methods for aligning proteins that have low sequence similarity to previously annotated proteins. We develop two deep learning methods to address this gap, TM-Vec and DeepBLAST. TM-Vec allows searching for structure-structure similarities in large sequence databases. It is trained to accurately predict TM-scores as a metric of structural similarity directly from sequence pairs without the need for intermediate computation or solution of structures. Once structurally similar proteins have been identified, DeepBLAST can structurally align proteins using only sequence information by identifying structurally homologous regions between proteins. It outperforms traditional sequence alignment methods and performs similarly to structure-based alignment methods. We show the merits of TM-Vec and DeepBLAST on a variety of datasets, including better identification of remotely homologous proteins compared with state-of-the-art sequence alignment and structure prediction methods.
Collapse
Grants
- R35GM122515 National Science Foundation (NSF)
- IOS-1546218 National Science Foundation (NSF)
- R35 GM122515 NIGMS NIH HHS
- R01 DK103358 NIDDK NIH HHS
- CBET- 1728858 National Science Foundation (NSF)
- R01 AI130945 NIAID NIH HHS
- This research was supported by NIH R01DK103358, the Simons Foundation, NSF- IOS-1546218, R35GM122515, NSF CBET- 1728858, NIH R01AI130945, to T.H. This research was supported by the intramural research program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) to J.T.M. This research was supported by the Flatiron Institute as part of the Simons Foundation to Robert Blackwell, J.K.L., and N.C. This research was supported by Los Alamos National Lab to C.S. This research was supported by the Samsung Advanced Institute of Technology (Next Generation Deep Learning: from pattern recognition to AI), Samsung Research (Improving Deep Learning using Latent Structure), and NSF Award 1922658 to K.C.
- Simons Foundation
- U.S. Department of Health & Human Services | NIH | Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD)
Collapse
Affiliation(s)
- Tymor Hamamsy
- Center for Data Science, New York University, New York, NY, USA
| | - James T Morton
- Center for Computational Biology, Flatiron Institute, Simons Foundation, New York, NY, USA
- Biostatistics and Bioinformatics Branch, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA
| | - Robert Blackwell
- Scientific Computing Core, Flatiron Institute, Simons Foundation, New York, NY, USA
| | - Daniel Berenberg
- Department of Computer Science, Courant Institute of Mathematical Sciences, New York University, New York, NY, USA
- Prescient Design, New York, NY, USA
| | - Nicholas Carriero
- Scientific Computing Core, Flatiron Institute, Simons Foundation, New York, NY, USA
| | | | | | - Julia Koehler Leman
- Center for Computational Biology, Flatiron Institute, Simons Foundation, New York, NY, USA
| | - Kyunghyun Cho
- Center for Data Science, New York University, New York, NY, USA.
- Department of Computer Science, Courant Institute of Mathematical Sciences, New York University, New York, NY, USA.
- Prescient Design, New York, NY, USA.
- CIFAR, Toronto, Ontario, Canada.
| | - Richard Bonneau
- Center for Data Science, New York University, New York, NY, USA.
- Department of Computer Science, Courant Institute of Mathematical Sciences, New York University, New York, NY, USA.
- Prescient Design, New York, NY, USA.
- Department of Biology, New York University, New York, NY, USA.
| |
Collapse
|
7
|
Ding K, Luo J, Luo Y. Leveraging conformal prediction to annotate enzyme function space with limited false positives. PLoS Comput Biol 2024; 20:e1012135. [PMID: 38809942 PMCID: PMC11164347 DOI: 10.1371/journal.pcbi.1012135] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/02/2023] [Revised: 06/10/2024] [Accepted: 05/03/2024] [Indexed: 05/31/2024] Open
Abstract
Machine learning (ML) is increasingly being used to guide biological discovery in biomedicine such as prioritizing promising small molecules in drug discovery. In those applications, ML models are used to predict the properties of biological systems, and researchers use these predictions to prioritize candidates as new biological hypotheses for downstream experimental validations. However, when applied to unseen situations, these models can be overconfident and produce a large number of false positives. One solution to address this issue is to quantify the model's prediction uncertainty and provide a set of hypotheses with a controlled false discovery rate (FDR) pre-specified by researchers. We propose CPEC, an ML framework for FDR-controlled biological discovery. We demonstrate its effectiveness using enzyme function annotation as a case study, simulating the discovery process of identifying the functions of less-characterized enzymes. CPEC integrates a deep learning model with a statistical tool known as conformal prediction, providing accurate and FDR-controlled function predictions for a given protein enzyme. Conformal prediction provides rigorous statistical guarantees to the predictive model and ensures that the expected FDR will not exceed a user-specified level with high probability. Evaluation experiments show that CPEC achieves reliable FDR control, better or comparable prediction performance at a lower FDR than existing methods, and accurate predictions for enzymes under-represented in the training data. We expect CPEC to be a useful tool for biological discovery applications where a high yield rate in validation experiments is desired but the experimental budget is limited.
Collapse
Affiliation(s)
- Kerr Ding
- School of Computational Science and Engineering, Georgia Institute of Technology, Atlanta, Georgia, United States of America
| | - Jiaqi Luo
- School of Computational Science and Engineering, Georgia Institute of Technology, Atlanta, Georgia, United States of America
| | - Yunan Luo
- School of Computational Science and Engineering, Georgia Institute of Technology, Atlanta, Georgia, United States of America
| |
Collapse
|
8
|
Susanty M, Naim Mursalim MK, Hertadi R, Purwarianti A, Rajab TLE. Classifying alkaliphilic proteins using embeddings from protein language model. Comput Biol Med 2024; 173:108385. [PMID: 38547659 DOI: 10.1016/j.compbiomed.2024.108385] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/13/2023] [Revised: 03/22/2024] [Accepted: 03/24/2024] [Indexed: 04/17/2024]
Abstract
Alkaliphilic proteins have great potential as biocatalysts in biotechnology, especially for enzyme engineering. Extensive research has focused on exploring the enzymatic potential of alkaliphiles and characterizing alkaliphilic proteins. However, the current method employed for identifying these proteins that requires web lab experiment is time-consuming, labor-intensive, and expensive. Therefore, the development of a computational method for alkaliphilic protein identification would be invaluable for protein engineering and design. In this study, we present a novel approach that uses embeddings from a protein language model called ESM-2(3B) in a deep learning framework to classify alkaliphilic and non-alkaliphilic proteins. To our knowledge, this is the first attempt to employ embeddings from a pre-trained protein language model to classify alkaliphilic protein. A reliable dataset comprising 1,002 alkaliphilic and 1,866 non-alkaliphilic proteins was constructed for training and testing the proposed model. The proposed model, dubbed ALPACA, achieves performance scores of 0.88, 0.84, and 0.75 for accuracy, f1-score, and Matthew correlation coefficient respectively on independent dataset. ALPACA is likely to serve as a valuable resource for exploring protein alkalinity and its role in protein design and engineering.
Collapse
Affiliation(s)
- Meredita Susanty
- Institut Teknologi Bandung School of Electrical Engineering and Informatics, Jl. Ganesa 10, Bandung, Jawa Barat, Indonesia; Universitas Pertamina, School of Computer Science, Jl Teuku Nyak Arief Jakarta Selatan DKI Jakarta, Indonesia
| | - Muhammad Khaerul Naim Mursalim
- Institut Teknologi Bandung School of Electrical Engineering and Informatics, Jl. Ganesa 10, Bandung, Jawa Barat, Indonesia; Universitas Universal, Kompleks Maha Vihara Duta Maitreya Bukit Beruntung, Sei Panas Batam, 29456, Kepulauan Riau, Indonesia
| | - Rukman Hertadi
- Institut Teknologi Bandung Faculty of Math and Natural Sciences, Jl. Ganesa 10, Bandung, Jawa Barat, Indonesia
| | - Ayu Purwarianti
- Institut Teknologi Bandung School of Electrical Engineering and Informatics, Jl. Ganesa 10, Bandung, Jawa Barat, Indonesia; Center for Artificial Intelligence (U-CoE AI-VLB), Institut Teknologi Bandung, Bandung, Indonesia
| | - Tati LE Rajab
- Institut Teknologi Bandung School of Electrical Engineering and Informatics, Jl. Ganesa 10, Bandung, Jawa Barat, Indonesia.
| |
Collapse
|
9
|
Li C, Stražar M, Mohamed AMT, Pacheco JA, Walker RL, Lebar T, Zhao S, Lockart J, Dame A, Thurimella K, Jeanfavre S, Brown EM, Ang QY, Berdy B, Sergio D, Invernizzi R, Tinoco A, Pishchany G, Vasan RS, Balskus E, Huttenhower C, Vlamakis H, Clish C, Shaw SY, Plichta DR, Xavier RJ. Gut microbiome and metabolome profiling in Framingham heart study reveals cholesterol-metabolizing bacteria. Cell 2024; 187:1834-1852.e19. [PMID: 38569543 PMCID: PMC11071153 DOI: 10.1016/j.cell.2024.03.014] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/06/2023] [Revised: 01/23/2024] [Accepted: 03/11/2024] [Indexed: 04/05/2024]
Abstract
Accumulating evidence suggests that cardiovascular disease (CVD) is associated with an altered gut microbiome. Our understanding of the underlying mechanisms has been hindered by lack of matched multi-omic data with diagnostic biomarkers. To comprehensively profile gut microbiome contributions to CVD, we generated stool metagenomics and metabolomics from 1,429 Framingham Heart Study participants. We identified blood lipids and cardiovascular health measurements associated with microbiome and metabolome composition. Integrated analysis revealed microbial pathways implicated in CVD, including flavonoid, γ-butyrobetaine, and cholesterol metabolism. Species from the Oscillibacter genus were associated with decreased fecal and plasma cholesterol levels. Using functional prediction and in vitro characterization of multiple representative human gut Oscillibacter isolates, we uncovered conserved cholesterol-metabolizing capabilities, including glycosylation and dehydrogenation. These findings suggest that cholesterol metabolism is a broad property of phylogenetically diverse Oscillibacter spp., with potential benefits for lipid homeostasis and cardiovascular health.
Collapse
Affiliation(s)
- Chenhao Li
- Broad Institute of MIT and Harvard, Cambridge, MA, USA; Center for Computational and Integrative Biology and Department of Molecular Biology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| | | | - Ahmed M T Mohamed
- Broad Institute of MIT and Harvard, Cambridge, MA, USA; Center for Computational and Integrative Biology and Department of Molecular Biology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| | | | | | - Tina Lebar
- Wyss Institute for Biologically Inspired Engineering, Harvard University, Boston, MA 02115, USA
| | - Shijie Zhao
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
| | - Julia Lockart
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
| | - Andrea Dame
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
| | | | | | - Eric M Brown
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
| | - Qi Yan Ang
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
| | | | - Dallis Sergio
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
| | - Rachele Invernizzi
- Broad Institute of MIT and Harvard, Cambridge, MA, USA; Center for Computational and Integrative Biology and Department of Molecular Biology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| | - Antonio Tinoco
- Department of Chemistry and Chemical Biology, Harvard University, Cambridge, MA, USA
| | | | - Ramachandran S Vasan
- Boston University and NHLBI's Framingham Heart Study, Framingham, MA, USA; Sections of Preventive Medicine and Epidemiology and Cardiology, Department of Medicine, Boston University School of Medicine, Boston, MA, USA; University of Texas School of Public Health, San Antonio, TX, USA
| | - Emily Balskus
- Department of Chemistry and Chemical Biology, Harvard University, Cambridge, MA, USA; Howard Hughes Medical Institute, Harvard University, Cambridge, MA, USA
| | - Curtis Huttenhower
- Broad Institute of MIT and Harvard, Cambridge, MA, USA; Harvard T.H. Chan School of Public Health, Boston, MA, USA
| | - Hera Vlamakis
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
| | - Clary Clish
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
| | - Stanley Y Shaw
- Broad Institute of MIT and Harvard, Cambridge, MA, USA; Division of Cardiovascular Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA
| | | | - Ramnik J Xavier
- Broad Institute of MIT and Harvard, Cambridge, MA, USA; Center for Computational and Integrative Biology and Department of Molecular Biology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA.
| |
Collapse
|
10
|
Liu W, Wang Z, You R, Xie C, Wei H, Xiong Y, Yang J, Zhu S. PLMSearch: Protein language model powers accurate and fast sequence search for remote homology. Nat Commun 2024; 15:2775. [PMID: 38555371 PMCID: PMC10981738 DOI: 10.1038/s41467-024-46808-5] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/28/2023] [Accepted: 03/08/2024] [Indexed: 04/02/2024] Open
Abstract
Homologous protein search is one of the most commonly used methods for protein annotation and analysis. Compared to structure search, detecting distant evolutionary relationships from sequences alone remains challenging. Here we propose PLMSearch (Protein Language Model), a homologous protein search method with only sequences as input. PLMSearch uses deep representations from a pre-trained protein language model and trains the similarity prediction model with a large number of real structure similarity. This enables PLMSearch to capture the remote homology information concealed behind the sequences. Extensive experimental results show that PLMSearch can search millions of query-target protein pairs in seconds like MMseqs2 while increasing the sensitivity by more than threefold, and is comparable to state-of-the-art structure search methods. In particular, unlike traditional sequence search methods, PLMSearch can recall most remote homology pairs with dissimilar sequences but similar structures. PLMSearch is freely available at https://dmiip.sjtu.edu.cn/PLMSearch .
Collapse
Affiliation(s)
- Wei Liu
- Institute of Science and Technology for Brain-Inspired Intelligence and MOE Frontiers Center for Brain Science, Fudan University, 200433, Shanghai, China
| | - Ziye Wang
- Institute of Science and Technology for Brain-Inspired Intelligence and MOE Frontiers Center for Brain Science, Fudan University, 200433, Shanghai, China
| | - Ronghui You
- Institute of Science and Technology for Brain-Inspired Intelligence and MOE Frontiers Center for Brain Science, Fudan University, 200433, Shanghai, China
| | - Chenghan Xie
- School of Mathematical Sciences, Fudan University, 200433, Shanghai, China
| | - Hong Wei
- School of Mathematical Sciences, Nankai University, 300071, Tianjin, China
| | - Yi Xiong
- Department of Bioinformatics and Biostatistics, Shanghai Jiao Tong University, 200240, Shanghai, China
| | - Jianyi Yang
- Ministry of Education Frontiers Science Center for Nonlinear Expectations, Research Center for Mathematics and Interdisciplinary Science, Shandong University, 266237, Qingdao, China.
| | - Shanfeng Zhu
- Institute of Science and Technology for Brain-Inspired Intelligence and MOE Frontiers Center for Brain Science, Fudan University, 200433, Shanghai, China.
- Shanghai Qi Zhi Institute, Shanghai, China.
- Key Laboratory of Computational Neuroscience and Brain-Inspired Intelligence (Fudan University), Ministry of Education, Shanghai, China.
- Shanghai Key Lab of Intelligent Information Processing and Shanghai Institute of Artificial Intelligence Algorithm, Fudan University, Shanghai, China.
- Zhangjiang Fudan International Innovation Center, Shanghai, China.
| |
Collapse
|
11
|
Lee B, Shin D. Contrastive learning for enhancing feature extraction in anticancer peptides. Brief Bioinform 2024; 25:bbae220. [PMID: 38725157 PMCID: PMC11082072 DOI: 10.1093/bib/bbae220] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/08/2024] [Revised: 03/28/2024] [Accepted: 04/21/2024] [Indexed: 05/13/2024] Open
Abstract
Cancer, recognized as a primary cause of death worldwide, has profound health implications and incurs a substantial social burden. Numerous efforts have been made to develop cancer treatments, among which anticancer peptides (ACPs) are garnering recognition for their potential applications. While ACP screening is time-consuming and costly, in silico prediction tools provide a way to overcome these challenges. Herein, we present a deep learning model designed to screen ACPs using peptide sequences only. A contrastive learning technique was applied to enhance model performance, yielding better results than a model trained solely on binary classification loss. Furthermore, two independent encoders were employed as a replacement for data augmentation, a technique commonly used in contrastive learning. Our model achieved superior performance on five of six benchmark datasets against previous state-of-the-art models. As prediction tools advance, the potential in peptide-based cancer therapeutics increases, promising a brighter future for oncology research and patient care.
Collapse
Affiliation(s)
- Byungjo Lee
- Research Institute, National Cancer Center, 323, Ilsan-ro, Ilsandong-gu, Goyang, 10408, Republic of Korea
| | - Dongkwan Shin
- Research Institute, National Cancer Center, 323, Ilsan-ro, Ilsandong-gu, Goyang, 10408, Republic of Korea
- Department of Cancer Biomedical Science, National Cancer Center Graduate School of Cancer Science and Policy, 323, Ilsan-ro, Ilsandong-gu, Goyang, 10408, Republic of Korea
| |
Collapse
|
12
|
Meng L, Chen X, Cheng K, Chen N, Zheng Z, Wang F, Sun H, Wong KC. TransPTM: a transformer-based model for non-histone acetylation site prediction. Brief Bioinform 2024; 25:bbae219. [PMID: 38725156 PMCID: PMC11082075 DOI: 10.1093/bib/bbae219] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/03/2023] [Revised: 04/08/2024] [Accepted: 04/23/2024] [Indexed: 05/13/2024] Open
Abstract
Protein acetylation is one of the extensively studied post-translational modifications (PTMs) due to its significant roles across a myriad of biological processes. Although many computational tools for acetylation site identification have been developed, there is a lack of benchmark dataset and bespoke predictors for non-histone acetylation site prediction. To address these problems, we have contributed to both dataset creation and predictor benchmark in this study. First, we construct a non-histone acetylation site benchmark dataset, namely NHAC, which includes 11 subsets according to the sequence length ranging from 11 to 61 amino acids. There are totally 886 positive samples and 4707 negative samples for each sequence length. Secondly, we propose TransPTM, a transformer-based neural network model for non-histone acetylation site predication. During the data representation phase, per-residue contextualized embeddings are extracted using ProtT5 (an existing pre-trained protein language model). This is followed by the implementation of a graph neural network framework, which consists of three TransformerConv layers for feature extraction and a multilayer perceptron module for classification. The benchmark results reflect that TransPTM has the competitive performance for non-histone acetylation site prediction over three state-of-the-art tools. It improves our comprehension on the PTM mechanism and provides a theoretical basis for developing drug targets for diseases. Moreover, the created PTM datasets fills the gap in non-histone acetylation site datasets and is beneficial to the related communities. The related source code and data utilized by TransPTM are accessible at https://www.github.com/TransPTM/TransPTM.
Collapse
Affiliation(s)
- Lingkuan Meng
- Department of Computer Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong
| | - Xingjian Chen
- Cutaneous Biology Research Center, Massachusetts General Hospital, Harvard Medical School, MA 02138, United States
| | - Ke Cheng
- Department of Chemistry, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong
| | - Nanjun Chen
- Department of Computer Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong
| | - Zetian Zheng
- Department of Computer Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong
| | - Fuzhou Wang
- Department of Computer Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong
| | - Hongyan Sun
- Department of Chemistry, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong
| | - Ka-Chun Wong
- Department of Computer Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong
- Shenzhen Research Institute, City University of Hong Kong, Shenzhen, China
| |
Collapse
|
13
|
Palacios A, Acharya P, Peidl A, Beck M, Blanco E, Mishra A, Bawa-Khalfe T, Pakhrin S. SumoPred-PLM: human SUMOylation and SUMO2/3 sites Prediction using Pre-trained Protein Language Model. NAR Genom Bioinform 2024; 6:lqae011. [PMID: 38327870 PMCID: PMC10849187 DOI: 10.1093/nargab/lqae011] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/21/2023] [Revised: 11/17/2023] [Accepted: 01/17/2024] [Indexed: 02/09/2024] Open
Abstract
SUMOylation is an essential post-translational modification system with the ability to regulate nearly all aspects of cellular physiology. Three major paralogues SUMO1, SUMO2 and SUMO3 form a covalent bond between the small ubiquitin-like modifier with lysine residues at consensus sites in protein substrates. Biochemical studies continue to identify unique biological functions for protein targets conjugated to SUMO1 versus the highly homologous SUMO2 and SUMO3 paralogues. Yet, the field has failed to harness contemporary AI approaches including pre-trained protein language models to fully expand and/or recognize the SUMOylated proteome. Herein, we present a novel, deep learning-based approach called SumoPred-PLM for human SUMOylation prediction with sensitivity, specificity, Matthew's correlation coefficient, and accuracy of 74.64%, 73.36%, 0.48% and 74.00%, respectively, on the CPLM 4.0 independent test dataset. In addition, this novel platform uses contextualized embeddings obtained from a pre-trained protein language model, ProtT5-XL-UniRef50 to identify SUMO2/3-specific conjugation sites. The results demonstrate that SumoPred-PLM is a powerful and unique computational tool to predict SUMOylation sites in proteins and accelerate discovery.
Collapse
Affiliation(s)
- Andrew Vargas Palacios
- Department of Computer Science and Engineering Technology, University of Houston-Downtown, 1 Main St., Houston, TX 77002, USA
| | - Pujan Acharya
- Department of Computer Science and Engineering Technology, University of Houston-Downtown, 1 Main St., Houston, TX 77002, USA
| | - Anthony Stephen Peidl
- Department of Biology and Biochemistry, Center for Nuclear Receptors & Cell Signaling, University of Houston, Houston, TX 77204, USA
| | - Moriah Rene Beck
- Department of Chemistry and Biochemistry, Wichita State University, 1845 Fairmount St., Wichita, KS 67260, USA
| | - Eduardo Blanco
- Department of Computer Science, University of Arizona, 1040 4th St., Tucson, AZ 85721, USA
| | - Avdesh Mishra
- Department of Electrical Engineering and Computer Science, Texas A&M University-Kingsville, Kingsville, TX 78363, USA
| | - Tasneem Bawa-Khalfe
- Department of Biology and Biochemistry, Center for Nuclear Receptors & Cell Signaling, University of Houston, Houston, TX 77204, USA
| | - Subash Chandra Pakhrin
- Department of Computer Science and Engineering Technology, University of Houston-Downtown, 1 Main St., Houston, TX 77002, USA
| |
Collapse
|
14
|
Pantolini L, Studer G, Pereira J, Durairaj J, Tauriello G, Schwede T. Embedding-based alignment: combining protein language models with dynamic programming alignment to detect structural similarities in the twilight-zone. Bioinformatics 2024; 40:btad786. [PMID: 38175775 PMCID: PMC10792726 DOI: 10.1093/bioinformatics/btad786] [Citation(s) in RCA: 4] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/10/2023] [Revised: 10/27/2023] [Accepted: 12/29/2023] [Indexed: 01/06/2024] Open
Abstract
MOTIVATION Language models are routinely used for text classification and generative tasks. Recently, the same architectures were applied to protein sequences, unlocking powerful new approaches in the bioinformatics field. Protein language models (pLMs) generate high-dimensional embeddings on a per-residue level and encode a "semantic meaning" of each individual amino acid in the context of the full protein sequence. These representations have been used as a starting point for downstream learning tasks and, more recently, for identifying distant homologous relationships between proteins. RESULTS In this work, we introduce a new method that generates embedding-based protein sequence alignments (EBA) and show how these capture structural similarities even in the twilight zone, outperforming both classical methods as well as other approaches based on pLMs. The method shows excellent accuracy despite the absence of training and parameter optimization. We demonstrate that the combination of pLMs with alignment methods is a valuable approach for the detection of relationships between proteins in the twilight-zone. AVAILABILITY AND IMPLEMENTATION The code to run EBA and reproduce the analysis described in this article is available at: https://git.scicore.unibas.ch/schwede/EBA and https://git.scicore.unibas.ch/schwede/eba_benchmark.
Collapse
Affiliation(s)
- Lorenzo Pantolini
- Biozentrum, University of Basel, Basel 4056, Switzerland
- SIB Swiss Institute of Bioinformatics, Basel 4056, Switzerland
| | - Gabriel Studer
- Biozentrum, University of Basel, Basel 4056, Switzerland
- SIB Swiss Institute of Bioinformatics, Basel 4056, Switzerland
| | - Joana Pereira
- Biozentrum, University of Basel, Basel 4056, Switzerland
- SIB Swiss Institute of Bioinformatics, Basel 4056, Switzerland
| | - Janani Durairaj
- Biozentrum, University of Basel, Basel 4056, Switzerland
- SIB Swiss Institute of Bioinformatics, Basel 4056, Switzerland
| | - Gerardo Tauriello
- Biozentrum, University of Basel, Basel 4056, Switzerland
- SIB Swiss Institute of Bioinformatics, Basel 4056, Switzerland
| | - Torsten Schwede
- Biozentrum, University of Basel, Basel 4056, Switzerland
- SIB Swiss Institute of Bioinformatics, Basel 4056, Switzerland
| |
Collapse
|
15
|
Urhan A, Cosma BM, Earl AM, Manson AL, Abeel T. SAP: Synteny-aware gene function prediction for bacteria using protein embeddings. BIORXIV : THE PREPRINT SERVER FOR BIOLOGY 2023:2023.05.02.539034. [PMID: 37205418 PMCID: PMC10187222 DOI: 10.1101/2023.05.02.539034] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/21/2023]
Abstract
Motivation Today, we know the function of only a small fraction of the protein sequences predicted from genomic data. This problem is even more salient for bacteria, which represent some of the most phylogenetically and metabolically diverse taxa on Earth. This low rate of bacterial gene annotation is compounded by the fact that most function prediction algorithms have focused on eukaryotes, and conventional annotation approaches rely on the presence of similar sequences in existing databases. However, often there are no such sequences for novel bacterial proteins. Thus, we need improved gene function prediction methods tailored for prokaryotes. Recently, transformer-based language models - adopted from the natural language processing field - have been used to obtain new representations of proteins, to replace amino acid sequences. These representations, referred to as protein embeddings, have shown promise for improving annotation of eukaryotes, but there have been only limited applications on bacterial genomes. Results To predict gene functions in bacteria, we developed SAP, a novel synteny-aware gene function prediction tool based on protein embeddings from state-of-the-art protein language models. SAP also leverages the unique operon structure of bacteria through conserved synteny. SAP outperformed both conventional sequence-based annotation methods and state-of-the-art methods on multiple bacterial species, including for distant homolog detection, where the sequence similarity to the proteins in the training set was as low as 40%. Using SAP to identify gene functions across diverse enterococci, of which some species are major clinical threats, we identified 11 previously unrecognized putative novel toxins, with potential significance to human and animal health.
Collapse
Affiliation(s)
- Aysun Urhan
- Delft Bioinformatics Lab, Delft University of Technology Van Mourik, Broekmanweg 6, 2628 XE, Delft, The Netherlands
- Infectious Disease and Microbiome Program, Broad Institute of MIT and Harvard, 415 Main Street, Cambridge, MA, 02142, US
| | - Bianca-Maria Cosma
- Delft Bioinformatics Lab, Delft University of Technology Van Mourik, Broekmanweg 6, 2628 XE, Delft, The Netherlands
| | - Ashlee M. Earl
- Infectious Disease and Microbiome Program, Broad Institute of MIT and Harvard, 415 Main Street, Cambridge, MA, 02142, US
| | - Abigail L. Manson
- Infectious Disease and Microbiome Program, Broad Institute of MIT and Harvard, 415 Main Street, Cambridge, MA, 02142, US
| | - Thomas Abeel
- Delft Bioinformatics Lab, Delft University of Technology Van Mourik, Broekmanweg 6, 2628 XE, Delft, The Netherlands
- Infectious Disease and Microbiome Program, Broad Institute of MIT and Harvard, 415 Main Street, Cambridge, MA, 02142, US
| |
Collapse
|
16
|
Thurimella K, Mohamed AMT, Graham DB, Owens RM, La Rosa SL, Plichta DR, Bacallado S, Xavier RJ. Protein Language Models Uncover Carbohydrate-Active Enzyme Function in Metagenomics. BIORXIV : THE PREPRINT SERVER FOR BIOLOGY 2023:2023.10.23.563620. [PMID: 37961379 PMCID: PMC10634757 DOI: 10.1101/2023.10.23.563620] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/15/2023]
Abstract
In metagenomics, the pool of uncharacterized microbial enzymes presents a challenge for functional annotation. Among these, carbohydrate-active enzymes (CAZymes) stand out due to their pivotal roles in various biological processes related to host health and nutrition. Here, we present CAZyLingua, the first tool that harnesses protein language model embeddings to build a deep learning framework that facilitates the annotation of CAZymes in metagenomic datasets. Our benchmarking results showed on average a higher F1 score (reflecting an average of precision and recall) on the annotated genomes of Bacteroides thetaiotaomicron, Eggerthella lenta and Ruminococcus gnavus compared to the traditional sequence homology-based method in dbCAN2. We applied our tool to a paired mother/infant longitudinal dataset and revealed unannotated CAZymes linked to microbial development during infancy. When applied to metagenomic datasets derived from patients affected by fibrosis-prone diseases such as Crohn's disease and IgG4-related disease, CAZyLingua uncovered CAZymes associated with disease and healthy states. In each of these metagenomic catalogs, CAZyLingua discovered new annotations that were previously overlooked by traditional sequence homology tools. Overall, the deep learning model CAZyLingua can be applied in combination with existing tools to unravel intricate CAZyme evolutionary profiles and patterns, contributing to a more comprehensive understanding of microbial metabolic dynamics.
Collapse
Affiliation(s)
- Kumar Thurimella
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
- Center for Computational and Integrative Biology and Department of Molecular Biology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
- Department of Chemical Engineering and Biotechnology, University of Cambridge, Cambridge, UK
- School of Medicine, University of Colorado Anschutz Medical Campus, Aurora, CO, USA
| | - Ahmed M. T. Mohamed
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
- Center for Computational and Integrative Biology and Department of Molecular Biology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| | - Daniel B. Graham
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
- Center for Computational and Integrative Biology and Department of Molecular Biology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| | - Róisín M. Owens
- Department of Chemical Engineering and Biotechnology, University of Cambridge, Cambridge, UK
| | - Sabina Leanti La Rosa
- Faculty of Chemistry, Biotechnology and Food Science, Norwegian University of Life Sciences, Ås, Norway
| | - Damian R. Plichta
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
- Center for Computational and Integrative Biology and Department of Molecular Biology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| | - Sergio Bacallado
- Department of Pure Mathematics and Mathematical Statistics, University of Cambridge, Cambridge, UK
| | - Ramnik J. Xavier
- Broad Institute of MIT and Harvard, Cambridge, MA, USA
- Center for Computational and Integrative Biology and Department of Molecular Biology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| |
Collapse
|
17
|
Koludarov I, Senoner T, Jackson TNW, Dashevsky D, Heinzinger M, Aird SD, Rost B. Domain loss enabled evolution of novel functions in the snake three-finger toxin gene superfamily. Nat Commun 2023; 14:4861. [PMID: 37567881 PMCID: PMC10421932 DOI: 10.1038/s41467-023-40550-0] [Citation(s) in RCA: 5] [Impact Index Per Article: 5.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/15/2022] [Accepted: 07/28/2023] [Indexed: 08/13/2023] Open
Abstract
Three-finger toxins (3FTXs) are a functionally diverse family of toxins, apparently unique to venoms of caenophidian snakes. Although the ancestral function of 3FTXs is antagonism of nicotinic acetylcholine receptors, redundancy conferred by the accumulation of duplicate genes has facilitated extensive neofunctionalization, such that derived members of the family interact with a range of targets. 3FTXs are members of the LY6/UPAR family, but their non-toxin ancestor remains unknown. Combining traditional phylogenetic approaches, manual synteny analysis, and machine learning techniques (including AlphaFold2 and ProtT5), we have reconstructed a detailed evolutionary history of 3FTXs. We identify their immediate ancestor as a non-secretory LY6, unique to squamate reptiles, and propose that changes in molecular ecology resulting from loss of a membrane-anchoring domain and changes in gene expression, paved the way for the evolution of one of the most important families of snake toxins.
Collapse
Affiliation(s)
- Ivan Koludarov
- TUM (Technical University of Munich) Department of Informatics, Bioinformatics & Computational Biology-i12, Boltzmannstr. 3, 85748, Garching/Munich, Germany.
| | - Tobias Senoner
- TUM (Technical University of Munich) Department of Informatics, Bioinformatics & Computational Biology-i12, Boltzmannstr. 3, 85748, Garching/Munich, Germany
| | - Timothy N W Jackson
- Australian Venom Research Unit, Department of Biochemistry and Pharmacology, University of Melbourne, Melbourne, VIC, Australia
| | - Daniel Dashevsky
- Australian National Insect Collection, Commonwealth Scientific & Industrial Research Organisation, Canberra, ACT, Australia
| | - Michael Heinzinger
- TUM (Technical University of Munich) Department of Informatics, Bioinformatics & Computational Biology-i12, Boltzmannstr. 3, 85748, Garching/Munich, Germany
| | - Steven D Aird
- 7744-23 Hotaka Ariake, 399-8301, Azumino-shi, Nagano-ken, Japan
| | - Burkhard Rost
- TUM (Technical University of Munich) Department of Informatics, Bioinformatics & Computational Biology-i12, Boltzmannstr. 3, 85748, Garching/Munich, Germany
- Institute for Advanced Study (TUM-IAS), Lichtenbergstr. 2a, 85748, Garching/Munich, Germany
- TUM School of Life Sciences Weihenstephan (WZW), Alte Akademie 8, Freising, Germany
| |
Collapse
|
18
|
McWhite CD, Armour-Garb I, Singh M. Leveraging protein language models for accurate multiple sequence alignments. Genome Res 2023; 33:1145-1153. [PMID: 37414576 PMCID: PMC10538487 DOI: 10.1101/gr.277675.123] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/06/2023] [Accepted: 06/29/2023] [Indexed: 07/08/2023]
Abstract
Multiple sequence alignment (MSA) is a critical step in the study of protein sequence and function. Typically, MSA algorithms progressively align pairs of sequences and combine these alignments with the aid of a guide tree. These alignment algorithms use scoring systems based on substitution matrices to measure amino acid similarities. Although successful, standard methods struggle on sets of proteins with low sequence identity: the so-called twilight zone of protein alignment. For these difficult cases, another source of information is needed. Protein language models are a powerful new approach that leverages massive sequence data sets to produce high-dimensional contextual embeddings for each amino acid in a sequence. These embeddings have been shown to reflect physicochemical and higher-order structural and functional attributes of amino acids within proteins. Here, we present a novel approach to MSA, based on clustering and ordering amino acid contextual embeddings. Our method for aligning semantically consistent groups of proteins circumvents the need for many standard components of MSA algorithms, avoiding initial guide tree construction, intermediate pairwise alignments, gap penalties, and substitution matrices. The added information from contextual embeddings leads to higher accuracy alignments for structurally similar proteins with low amino-acid similarity. We anticipate that protein language models will become a fundamental component of the next generation of algorithms for generating MSAs.
Collapse
Affiliation(s)
- Claire D McWhite
- Lewis-Sigler Institute for Integrative Genomics, Princeton University, Princeton, New Jersey 08544, USA;
| | - Isabel Armour-Garb
- Lewis-Sigler Institute for Integrative Genomics, Princeton University, Princeton, New Jersey 08544, USA
- Department of Computer Science, Princeton University, Princeton, New Jersey 08544, USA
| | - Mona Singh
- Lewis-Sigler Institute for Integrative Genomics, Princeton University, Princeton, New Jersey 08544, USA;
- Department of Computer Science, Princeton University, Princeton, New Jersey 08544, USA
| |
Collapse
|
19
|
Singh R, Sledzieski S, Bryson B, Cowen L, Berger B. Contrastive learning in protein language space predicts interactions between drugs and protein targets. Proc Natl Acad Sci U S A 2023; 120:e2220778120. [PMID: 37289807 PMCID: PMC10268324 DOI: 10.1073/pnas.2220778120] [Citation(s) in RCA: 24] [Impact Index Per Article: 24.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/06/2022] [Accepted: 04/10/2023] [Indexed: 06/10/2023] Open
Abstract
Sequence-based prediction of drug-target interactions has the potential to accelerate drug discovery by complementing experimental screens. Such computational prediction needs to be generalizable and scalable while remaining sensitive to subtle variations in the inputs. However, current computational techniques fail to simultaneously meet these goals, often sacrificing performance of one to achieve the others. We develop a deep learning model, ConPLex, successfully leveraging the advances in pretrained protein language models ("PLex") and employing a protein-anchored contrastive coembedding ("Con") to outperform state-of-the-art approaches. ConPLex achieves high accuracy, broad adaptivity to unseen data, and specificity against decoy compounds. It makes predictions of binding based on the distance between learned representations, enabling predictions at the scale of massive compound libraries and the human proteome. Experimental testing of 19 kinase-drug interaction predictions validated 12 interactions, including four with subnanomolar affinity, plus a strongly binding EPHB1 inhibitor (KD = 1.3 nM). Furthermore, ConPLex embeddings are interpretable, which enables us to visualize the drug-target embedding space and use embeddings to characterize the function of human cell-surface proteins. We anticipate that ConPLex will facilitate efficient drug discovery by making highly sensitive in silico drug screening feasible at the genome scale. ConPLex is available open source at https://ConPLex.csail.mit.edu.
Collapse
Affiliation(s)
- Rohit Singh
- Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA02139
| | - Samuel Sledzieski
- Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA02139
| | - Bryan Bryson
- Ragon Institute of MGH, MIT and Harvard, Cambridge, MA02139
- Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge, MA02139
| | - Lenore Cowen
- Department of Computer Science, Tufts University, Medford, MA02155
| | - Bonnie Berger
- Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA02139
- Department of Mathematics, Massachusetts Institute of Technology, Cambridge, MA02139
| |
Collapse
|
20
|
Bordin N, Dallago C, Heinzinger M, Kim S, Littmann M, Rauer C, Steinegger M, Rost B, Orengo C. Novel machine learning approaches revolutionize protein knowledge. Trends Biochem Sci 2023; 48:345-359. [PMID: 36504138 PMCID: PMC10570143 DOI: 10.1016/j.tibs.2022.11.001] [Citation(s) in RCA: 20] [Impact Index Per Article: 20.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/14/2022] [Revised: 10/24/2022] [Accepted: 11/17/2022] [Indexed: 12/10/2022]
Abstract
Breakthrough methods in machine learning (ML), protein structure prediction, and novel ultrafast structural aligners are revolutionizing structural biology. Obtaining accurate models of proteins and annotating their functions on a large scale is no longer limited by time and resources. The most recent method to be top ranked by the Critical Assessment of Structure Prediction (CASP) assessment, AlphaFold 2 (AF2), is capable of building structural models with an accuracy comparable to that of experimental structures. Annotations of 3D models are keeping pace with the deposition of the structures due to advancements in protein language models (pLMs) and structural aligners that help validate these transferred annotations. In this review we describe how recent developments in ML for protein science are making large-scale structural bioinformatics available to the general scientific community.
Collapse
Affiliation(s)
- Nicola Bordin
- Institute of Structural and Molecular Biology, University College London, Gower St, WC1E 6BT London, UK
| | - Christian Dallago
- Technical University of Munich (TUM) Department of Informatics, Bioinformatics and Computational Biology - i12, Boltzmannstr. 3, 85748 Garching/Munich, Germany; VantAI, 151 W 42nd Street, New York, NY 10036, USA
| | - Michael Heinzinger
- Technical University of Munich (TUM) Department of Informatics, Bioinformatics and Computational Biology - i12, Boltzmannstr. 3, 85748 Garching/Munich, Germany; TUM Graduate School, Center of Doctoral Studies in Informatics and its Applications (CeDoSIA), Boltzmannstr. 11, 85748 Garching, Germany
| | - Stephanie Kim
- School of Biological Sciences, Seoul National University, Seoul, South Korea; Artificial Intelligence Institute, Seoul National University, Seoul, South Korea
| | - Maria Littmann
- Technical University of Munich (TUM) Department of Informatics, Bioinformatics and Computational Biology - i12, Boltzmannstr. 3, 85748 Garching/Munich, Germany
| | - Clemens Rauer
- Institute of Structural and Molecular Biology, University College London, Gower St, WC1E 6BT London, UK
| | - Martin Steinegger
- School of Biological Sciences, Seoul National University, Seoul, South Korea; Artificial Intelligence Institute, Seoul National University, Seoul, South Korea
| | - Burkhard Rost
- Technical University of Munich (TUM) Department of Informatics, Bioinformatics and Computational Biology - i12, Boltzmannstr. 3, 85748 Garching/Munich, Germany; Institute for Advanced Study (TUM-IAS), Lichtenbergstr. 2a, 85748 Garching/Munich, Germany; TUM School of Life Sciences Weihenstephan (TUM-WZW), Alte Akademie 8, Freising, Germany
| | - Christine Orengo
- Institute of Structural and Molecular Biology, University College London, Gower St, WC1E 6BT London, UK.
| |
Collapse
|
21
|
Yu T, Boob AG, Volk MJ, Liu X, Cui H, Zhao H. Machine learning-enabled retrobiosynthesis of molecules. Nat Catal 2023. [DOI: 10.1038/s41929-022-00909-w] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 02/18/2023]
|
22
|
AlphaFold2 reveals commonalities and novelties in protein structure space for 21 model organisms. Commun Biol 2023; 6:160. [PMID: 36755055 PMCID: PMC9908985 DOI: 10.1038/s42003-023-04488-9] [Citation(s) in RCA: 29] [Impact Index Per Article: 29.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/06/2022] [Accepted: 01/16/2023] [Indexed: 02/10/2023] Open
Abstract
Deep-learning (DL) methods like DeepMind's AlphaFold2 (AF2) have led to substantial improvements in protein structure prediction. We analyse confident AF2 models from 21 model organisms using a new classification protocol (CATH-Assign) which exploits novel DL methods for structural comparison and classification. Of ~370,000 confident models, 92% can be assigned to 3253 superfamilies in our CATH domain superfamily classification. The remaining cluster into 2367 putative novel superfamilies. Detailed manual analysis on 618 of these, having at least one human relative, reveal extremely remote homologies and further unusual features. Only 25 novel superfamilies could be confirmed. Although most models map to existing superfamilies, AF2 domains expand CATH by 67% and increases the number of unique 'global' folds by 36% and will provide valuable insights on structure function relationships. CATH-Assign will harness the huge expansion in structural data provided by DeepMind to rationalise evolutionary changes driving functional divergence.
Collapse
|
23
|
Pratyush P, Pokharel S, Saigo H, KC DB. pLMSNOSite: an ensemble-based approach for predicting protein S-nitrosylation sites by integrating supervised word embedding and embedding from pre-trained protein language model. BMC Bioinformatics 2023; 24:41. [PMID: 36755242 PMCID: PMC9909867 DOI: 10.1186/s12859-023-05164-9] [Citation(s) in RCA: 6] [Impact Index Per Article: 6.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/14/2022] [Accepted: 01/30/2023] [Indexed: 02/10/2023] Open
Abstract
BACKGROUND Protein S-nitrosylation (SNO) plays a key role in transferring nitric oxide-mediated signals in both animals and plants and has emerged as an important mechanism for regulating protein functions and cell signaling of all main classes of protein. It is involved in several biological processes including immune response, protein stability, transcription regulation, post translational regulation, DNA damage repair, redox regulation, and is an emerging paradigm of redox signaling for protection against oxidative stress. The development of robust computational tools to predict protein SNO sites would contribute to further interpretation of the pathological and physiological mechanisms of SNO. RESULTS Using an intermediate fusion-based stacked generalization approach, we integrated embeddings from supervised embedding layer and contextualized protein language model (ProtT5) and developed a tool called pLMSNOSite (protein language model-based SNO site predictor). On an independent test set of experimentally identified SNO sites, pLMSNOSite achieved values of 0.340, 0.735 and 0.773 for MCC, sensitivity and specificity respectively. These results show that pLMSNOSite performs better than the compared approaches for the prediction of S-nitrosylation sites. CONCLUSION Together, the experimental results suggest that pLMSNOSite achieves significant improvement in the prediction performance of S-nitrosylation sites and represents a robust computational approach for predicting protein S-nitrosylation sites. pLMSNOSite could be a useful resource for further elucidation of SNO and is publicly available at https://github.com/KCLabMTU/pLMSNOSite .
Collapse
Affiliation(s)
- Pawel Pratyush
- grid.259979.90000 0001 0663 5937Department of Computer Science, Michigan Technological University, Houghton, MI USA
| | - Suresh Pokharel
- grid.259979.90000 0001 0663 5937Department of Computer Science, Michigan Technological University, Houghton, MI USA
| | - Hiroto Saigo
- grid.177174.30000 0001 2242 4849Department of Electrical Engineering and Computer Science, Kyushu University, 744, Motooka, Nishi-Ku, 819-0395 Japan
| | - Dukka B. KC
- grid.259979.90000 0001 0663 5937Department of Computer Science, Michigan Technological University, Houghton, MI USA
| |
Collapse
|
24
|
Chandra A, Tünnermann L, Löfstedt T, Gratz R. Transformer-based deep learning for predicting protein properties in the life sciences. eLife 2023; 12:82819. [PMID: 36651724 PMCID: PMC9848389 DOI: 10.7554/elife.82819] [Citation(s) in RCA: 14] [Impact Index Per Article: 14.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/22/2022] [Accepted: 01/06/2023] [Indexed: 01/19/2023] Open
Abstract
Recent developments in deep learning, coupled with an increasing number of sequenced proteins, have led to a breakthrough in life science applications, in particular in protein property prediction. There is hope that deep learning can close the gap between the number of sequenced proteins and proteins with known properties based on lab experiments. Language models from the field of natural language processing have gained popularity for protein property predictions and have led to a new computational revolution in biology, where old prediction results are being improved regularly. Such models can learn useful multipurpose representations of proteins from large open repositories of protein sequences and can be used, for instance, to predict protein properties. The field of natural language processing is growing quickly because of developments in a class of models based on a particular model-the Transformer model. We review recent developments and the use of large-scale Transformer models in applications for predicting protein characteristics and how such models can be used to predict, for example, post-translational modifications. We review shortcomings of other deep learning models and explain how the Transformer models have quickly proven to be a very promising way to unravel information hidden in the sequences of amino acids.
Collapse
Affiliation(s)
- Abel Chandra
- Department of Computing Science, Umeå UniversityUmeåSweden
| | - Laura Tünnermann
- Umeå Plant Science Centre (UPSC), Department of Forest Genetics and Plant Physiology, Swedish University of Agricultural SciencesUmeåSweden
| | - Tommy Löfstedt
- Department of Computing Science, Umeå UniversityUmeåSweden
| | - Regina Gratz
- Umeå Plant Science Centre (UPSC), Department of Forest Genetics and Plant Physiology, Swedish University of Agricultural SciencesUmeåSweden
- Department of Forest Ecology and Management, Swedish University of Agricultural SciencesUmeåSweden
| |
Collapse
|
25
|
Ferruz N, Heinzinger M, Akdel M, Goncearenco A, Naef L, Dallago C. From sequence to function through structure: Deep learning for protein design. Comput Struct Biotechnol J 2022; 21:238-250. [PMID: 36544476 PMCID: PMC9755234 DOI: 10.1016/j.csbj.2022.11.014] [Citation(s) in RCA: 27] [Impact Index Per Article: 13.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/31/2022] [Revised: 11/05/2022] [Accepted: 11/05/2022] [Indexed: 11/20/2022] Open
Abstract
The process of designing biomolecules, in particular proteins, is witnessing a rapid change in available tooling and approaches, moving from design through physicochemical force fields, to producing plausible, complex sequences fast via end-to-end differentiable statistical models. To achieve conditional and controllable protein design, researchers at the interface of artificial intelligence and biology leverage advances in natural language processing (NLP) and computer vision techniques, coupled with advances in computing hardware to learn patterns from growing biological databases, curated annotations thereof, or both. Once learned, these patterns can be used to provide novel insights into mechanistic biology and the design of biomolecules. However, navigating and understanding the practical applications for the many recent protein design tools is complex. To facilitate this, we 1) document recent advances in deep learning (DL) assisted protein design from the last three years, 2) present a practical pipeline that allows to go from de novo-generated sequences to their predicted properties and web-powered visualization within minutes, and 3) leverage it to suggest a generated protein sequence which might be used to engineer a biosynthetic gene cluster to produce a molecular glue-like compound. Lastly, we discuss challenges and highlight opportunities for the protein design field.
Collapse
Key Words
- ADMM, Alternating Direction Method of Multipliers
- CNN, Convolutional Neural Network
- DL, Deep learning
- Deep learning
- Drug discovery
- FNN, fully-connected neural network
- GAN, Generative Adversarial Network
- GCN, Graph Convolutional Network
- GNN, Graph Neural Network
- GO, Gene Ontology
- GVP, Geometric Vector Perceptron
- LSTM, Long-Short Term Memory
- MLP, Multilayer Perceptron
- MSA, Multiple Sequence Alignment
- NLP, Natural Language Processing
- NSR, Natural Sequence Recovery
- Protein design
- Protein language models
- Protein prediction
- VAE, Variational Autoencoder
- pLM, protein Language Model
Collapse
Affiliation(s)
- Noelia Ferruz
- Institute of Informatics and Applications, University of Girona, Girona, Spain
- Department of Biochemistry, University of Bayreuth, Bayreuth, Germany
| | - Michael Heinzinger
- Department of Informatics, Bioinformatics & Computational Biology, Technische Universität München, 85748 Garching, Germany
| | - Mehmet Akdel
- VantAI, 151 W 42nd Street, New York, NY 10036, United States
| | | | - Luca Naef
- VantAI, 151 W 42nd Street, New York, NY 10036, United States
| | - Christian Dallago
- Department of Informatics, Bioinformatics & Computational Biology, Technische Universität München, 85748 Garching, Germany
- VantAI, 151 W 42nd Street, New York, NY 10036, United States
- NVIDIA DE GmbH, Einsteinstraße 172, 81677 München, Germany
| |
Collapse
|
26
|
Schütze K, Heinzinger M, Steinegger M, Rost B. Nearest neighbor search on embeddings rapidly identifies distant protein relations. FRONTIERS IN BIOINFORMATICS 2022; 2:1033775. [PMID: 36466147 PMCID: PMC9714024 DOI: 10.3389/fbinf.2022.1033775] [Citation(s) in RCA: 8] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/31/2022] [Accepted: 10/31/2022] [Indexed: 11/29/2023] Open
Abstract
Since 1992, all state-of-the-art methods for fast and sensitive identification of evolutionary, structural, and functional relations between proteins (also referred to as "homology detection") use sequences and sequence-profiles (PSSMs). Protein Language Models (pLMs) generalize sequences, possibly capturing the same constraints as PSSMs, e.g., through embeddings. Here, we explored how to use such embeddings for nearest neighbor searches to identify relations between protein pairs with diverged sequences (remote homology detection for levels of <20% pairwise sequence identity, PIDE). While this approach excelled for proteins with single domains, we demonstrated the current challenges applying this to multi-domain proteins and presented some ideas how to overcome existing limitations, in principle. We observed that sufficiently challenging data set separations were crucial to provide deeply relevant insights into the behavior of nearest neighbor search when applied to the protein embedding space, and made all our methods readily available for others.
Collapse
Affiliation(s)
- Konstantin Schütze
- TUM (Technical University of Munich) Department of Informatics, Bioinformatics & Computational Biology—i12, Munich, Germany
| | - Michael Heinzinger
- TUM (Technical University of Munich) Department of Informatics, Bioinformatics & Computational Biology—i12, Munich, Germany
- TUM Graduate School, Center of Doctoral Studies in Informatics and its Applications (CeDoSIA), Garching, Germany
| | - Martin Steinegger
- School of Biological Sciences, Seoul National University, Seoul, South Korea
- Artificial Intelligence Institute, Seoul National University, Seoul, South Korea
| | - Burkhard Rost
- TUM (Technical University of Munich) Department of Informatics, Bioinformatics & Computational Biology—i12, Munich, Germany
- Institute for Advanced Study (TUM-IAS), Germany & TUM School of Life Sciences Weihenstephan (WZW), Freising, Germany
| |
Collapse
|
27
|
Ilzhöfer D, Heinzinger M, Rost B. SETH predicts nuances of residue disorder from protein embeddings. FRONTIERS IN BIOINFORMATICS 2022; 2:1019597. [PMID: 36304335 PMCID: PMC9580958 DOI: 10.3389/fbinf.2022.1019597] [Citation(s) in RCA: 7] [Impact Index Per Article: 3.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/15/2022] [Accepted: 09/20/2022] [Indexed: 11/07/2022] Open
Abstract
Predictions for millions of protein three-dimensional structures are only a few clicks away since the release of AlphaFold2 results for UniProt. However, many proteins have so-called intrinsically disordered regions (IDRs) that do not adopt unique structures in isolation. These IDRs are associated with several diseases, including Alzheimer's Disease. We showed that three recent disorder measures of AlphaFold2 predictions (pLDDT, "experimentally resolved" prediction and "relative solvent accessibility") correlated to some extent with IDRs. However, expert methods predict IDRs more reliably by combining complex machine learning models with expert-crafted input features and evolutionary information from multiple sequence alignments (MSAs). MSAs are not always available, especially for IDRs, and are computationally expensive to generate, limiting the scalability of the associated tools. Here, we present the novel method SETH that predicts residue disorder from embeddings generated by the protein Language Model ProtT5, which explicitly only uses single sequences as input. Thereby, our method, relying on a relatively shallow convolutional neural network, outperformed much more complex solutions while being much faster, allowing to create predictions for the human proteome in about 1 hour on a consumer-grade PC with one NVIDIA GeForce RTX 3060. Trained on a continuous disorder scale (CheZOD scores), our method captured subtle variations in disorder, thereby providing important information beyond the binary classification of most methods. High performance paired with speed revealed that SETH's nuanced disorder predictions for entire proteomes capture aspects of the evolution of organisms. Additionally, SETH could also be used to filter out regions or proteins with probable low-quality AlphaFold2 3D structures to prioritize running the compute-intensive predictions for large data sets. SETH is freely publicly available at: https://github.com/Rostlab/SETH.
Collapse
Affiliation(s)
- Dagmar Ilzhöfer
- Faculty of Informatics, TUM (Technical University of Munich), Munich, Germany
| | - Michael Heinzinger
- Faculty of Informatics, TUM (Technical University of Munich), Munich, Germany
- Center of Doctoral Studies in Informatics and Its Applications (CeDoSIA), TUM Graduate School, Garching, Germany
| | - Burkhard Rost
- Faculty of Informatics, TUM (Technical University of Munich), Munich, Germany
- Institute for Advanced Study (TUM-IAS), TUM (Technical University of Munich), Garching, Germany
- TUM School of Life Sciences Weihenstephan (WZW), TUM (Technical University of Munich), Freising, Germany
| |
Collapse
|
28
|
Capel H, Weiler R, Dijkstra M, Vleugels R, Bloem P, Feenstra KA. ProteinGLUE multi-task benchmark suite for self-supervised protein modeling. Sci Rep 2022; 12:16047. [PMID: 36163232 PMCID: PMC9512797 DOI: 10.1038/s41598-022-19608-4] [Citation(s) in RCA: 2] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/17/2021] [Accepted: 08/31/2022] [Indexed: 11/09/2022] Open
Abstract
Self-supervised language modeling is a rapidly developing approach for the analysis of protein sequence data. However, work in this area is heterogeneous and diverse, making comparison of models and methods difficult. Moreover, models are often evaluated only on one or two downstream tasks, making it unclear whether the models capture generally useful properties. We introduce the ProteinGLUE benchmark for the evaluation of protein representations: a set of seven per-amino-acid tasks for evaluating learned protein representations. We also offer reference code, and we provide two baseline models with hyperparameters specifically trained for these benchmarks. Pre-training was done on two tasks, masked symbol prediction and next sentence prediction. We show that pre-training yields higher performance on a variety of downstream tasks such as secondary structure and protein interaction interface prediction, compared to no pre-training. However, the larger base model does not outperform the smaller medium model. We expect the ProteinGLUE benchmark dataset introduced here, together with the two baseline pre-trained models and their performance evaluations, to be of great value to the field of protein sequence-based property prediction. Availability: code and datasets from https://github.com/ibivu/protein-glue .
Collapse
Affiliation(s)
- Henriette Capel
- Informatics Institute, Vrije Universiteit, 1081 HV, Amsterdam, The Netherlands
| | - Robin Weiler
- Informatics Institute, Vrije Universiteit, 1081 HV, Amsterdam, The Netherlands
| | - Maurits Dijkstra
- Informatics Institute, Vrije Universiteit, 1081 HV, Amsterdam, The Netherlands
| | - Reinier Vleugels
- Informatics Institute, Vrije Universiteit, 1081 HV, Amsterdam, The Netherlands
| | - Peter Bloem
- Informatics Institute, Vrije Universiteit, 1081 HV, Amsterdam, The Netherlands
| | - K Anton Feenstra
- Informatics Institute, Vrije Universiteit, 1081 HV, Amsterdam, The Netherlands.
| |
Collapse
|
29
|
Bernhofer M, Rost B. TMbed: transmembrane proteins predicted through language model embeddings. BMC Bioinformatics 2022; 23:326. [PMID: 35941534 PMCID: PMC9358067 DOI: 10.1186/s12859-022-04873-x] [Citation(s) in RCA: 16] [Impact Index Per Article: 8.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/12/2022] [Accepted: 08/03/2022] [Indexed: 12/30/2022] Open
Abstract
BACKGROUND Despite the immense importance of transmembrane proteins (TMP) for molecular biology and medicine, experimental 3D structures for TMPs remain about 4-5 times underrepresented compared to non-TMPs. Today's top methods such as AlphaFold2 accurately predict 3D structures for many TMPs, but annotating transmembrane regions remains a limiting step for proteome-wide predictions. RESULTS Here, we present TMbed, a novel method inputting embeddings from protein Language Models (pLMs, here ProtT5), to predict for each residue one of four classes: transmembrane helix (TMH), transmembrane strand (TMB), signal peptide, or other. TMbed completes predictions for entire proteomes within hours on a single consumer-grade desktop machine at performance levels similar or better than methods, which are using evolutionary information from multiple sequence alignments (MSAs) of protein families. On the per-protein level, TMbed correctly identified 94 ± 8% of the beta barrel TMPs (53 of 57) and 98 ± 1% of the alpha helical TMPs (557 of 571) in a non-redundant data set, at false positive rates well below 1% (erred on 30 of 5654 non-membrane proteins). On the per-segment level, TMbed correctly placed, on average, 9 of 10 transmembrane segments within five residues of the experimental observation. Our method can handle sequences of up to 4200 residues on standard graphics cards used in desktop PCs (e.g., NVIDIA GeForce RTX 3060). CONCLUSIONS Based on embeddings from pLMs and two novel filters (Gaussian and Viterbi), TMbed predicts alpha helical and beta barrel TMPs at least as accurately as any other method but at lower false positive rates. Given the few false positives and its outstanding speed, TMbed might be ideal to sieve through millions of 3D structures soon to be predicted, e.g., by AlphaFold2.
Collapse
Affiliation(s)
- Michael Bernhofer
- Department of Informatics, Bioinformatics and Computational Biology ‑ i12, Technical University of Munich (TUM), Boltzmannstr. 3, 85748, Garching, Germany.
- TUM Graduate School, Center of Doctoral Studies in Informatics and its Applications (CeDoSIA), Boltzmannstr. 11, 85748, Garching, Germany.
| | - Burkhard Rost
- Department of Informatics, Bioinformatics and Computational Biology ‑ i12, Technical University of Munich (TUM), Boltzmannstr. 3, 85748, Garching, Germany
- Institute for Advanced Study (TUM-IAS), Lichtenbergstr. 2a, 85748, Garching, Germany
- TUM School of Life Sciences Weihenstephan (TUM-WZW), Alte Akademie 8, Freising, Germany
| |
Collapse
|