FINER: enhancing the prediction of tissue-specific functions of isoforms by refining isoform interaction networks

Abstract Annotating the functions of gene products is a mainstay in biology. A variety of databases have been established to record functional knowledge at the gene level. However, functional annotations at the isoform resolution are in great demand in many biological applications. Although critical information in biological processes such as protein–protein interactions (PPIs) is often used to study gene functions, it does not directly help differentiate the functions of isoforms, as the ‘proteins’ in the existing PPIs generally refer to ‘genes’. On the other hand, the prediction of isoform functions and prediction of isoform–isoform interactions, though inherently intertwined, have so far been treated as independent computational problems in the literature. Here, we present FINER, a unified framework to jointly predict isoform functions and refine PPIs from the gene level to the isoform level, enabling both tasks to benefit from each other. Extensive computational experiments on human tissue-specific data demonstrate that FINER is able to gain at least 5.16% in AUC and 15.1% in AUPRC for functional prediction across multiple tissues by refining noisy PPIs, resulting in significant improvement over the state-of-the-art methods. Some in-depth analyses reveal consistency between FINER’s predictions and the tissue specificity as well as subcellular localization of isoforms.

Download Full-text

An Integrated Prediction Method for Identifying Protein-Protein Interactions

Current Proteomics ◽

10.2174/1570164616666190306152318 ◽

2020 ◽

Vol 17 (4) ◽

pp. 271-286

Author(s):

Chang Xu ◽

Limin Jiang ◽

Zehua Zhang ◽

Xuyao Yu ◽

Renhai Chen ◽

...

Keyword(s):

Protein Interactions ◽

Feature Vector ◽

Prediction Method ◽

Feature Representation ◽

Protein Interaction Networks ◽

Learning Approach ◽

Biological Processes ◽

Integrated Learning ◽

Protein Protein Interactions ◽

Training Process

Background: Protein-Protein Interactions (PPIs) play a key role in various biological processes. Many methods have been developed to predict protein-protein interactions and protein interaction networks. However, many existing applications are limited, because of relying on a large number of homology proteins and interaction marks. Methods: In this paper, we propose a novel integrated learning approach (RF-Ada-DF) with the sequence-based feature representation, for identifying protein-protein interactions. Our method firstly constructs a sequence-based feature vector to represent each pair of proteins, viaMultivariate Mutual Information (MMI) and Normalized Moreau-Broto Autocorrelation (NMBAC). Then, we feed the 638- dimentional features into an integrated learning model for judging interaction pairs and non-interaction pairs. Furthermore, this integrated model embeds Random Forest in AdaBoost framework and turns weak classifiers into a single strong classifier. Meanwhile, we also employ double fault detection in order to suppress over-adaptation during the training process. Results: To evaluate the performance of our method, we conduct several comprehensive tests for PPIs prediction. On the H. pyloridataset, our method achieves 88.16% accuracy and 87.68% sensitivity, the accuracy of our method is increased by 0.57%. On the S. cerevisiaedataset, our method achieves 95.77% accuracy and 93.36% sensitivity, the accuracy of our method is increased by 0.76%. On the Humandataset, our method achieves 98.16% accuracy and 96.80% sensitivity, the accuracy of our method is increased by 0.6%. Experiments show that our method achieves better results than other outstanding methods for sequence-based PPIs prediction. The datasets and codes are available at https://github.com/guofei-tju/RF-Ada-DF.git.

Download Full-text

Entamoeba histolytica: Proteomics Bioinformatics Reveal Predictive Functions and Protein–Protein Interactions of Differentially Abundant Membrane and Cytosolic Proteins

Membranes ◽

10.3390/membranes11060376 ◽

2021 ◽

Vol 11 (6) ◽

pp. 376

Author(s):

Norhidayah Azmi ◽

Nurulhasanah Othman

Keyword(s):

Entamoeba Histolytica ◽

Protein Interactions ◽

Biological Process ◽

Protein Protein Interactions ◽

Differential Abundance ◽

Cytosolic Proteins ◽

Functional Annotations ◽

Predictive Functions ◽

Differentially Abundant Proteins ◽

Abundance Proteins

Amoebiasis is caused by Entamoeba histolytica and ranked second for parasitic diseases causing death after malaria. E. histolytica membrane and cytosolic proteins play important roles in the pathogenesis. Our previous study had shown several cytosolic proteins were found in the membrane fraction. Therefore, this study aimed to quantify the differential abundance of membrane and cytosolic proteins in membrane versus cytosolic fractions and analyze their predicted functions and interaction. Previous LC-ESI-MS/MS data were analyzed by PERSEUS software for the differentially abundant proteins, then they were classified into their functional annotations and the protein networks were summarized using PantherDB and STRiNG, respectively. The results showed 24 (44.4%) out of the 54 proteins that increased in abundance were membrane proteins and 30 were cytosolic proteins. Meanwhile, 45 cytosolic proteins were found to decrease in abundance. Functional analysis showed differential abundance proteins involved in the molecular function, biological process, and cellular component with 18.88%, 33.04% and, 48.07%, respectively. The STRiNG server predicted that the decreased abundance proteins had more protein–protein network interactions compared to increased abundance proteins. Overall, this study has confirmed the presence of the differentially abundant membrane and cytosolic proteins and provided the predictive functions and interactions between them.

Download Full-text

A Modified TurboID Approach Identifies Tissue-Specific Centriolar Components In C. elegans

10.1101/2021.12.20.473533 ◽

2021 ◽

Author(s):

Elisabeth Holzer ◽

Cornelia Rumpf-Kienzl ◽

Sebastian Falk ◽

Alexander Dammermann

Keyword(s):

Protein Interactions ◽

Affinity Purification ◽

Experimental Models ◽

Protein Protein Interactions ◽

Tissue Specific ◽

C Elegans ◽

Early Embryos ◽

Somatic Tissues ◽

Specific Variation ◽

Labeling Method

Proximity-dependent labeling approaches such as BioID have been a great boon to studies of protein-protein interactions in the context of cytoskeletal structures such as centrosomes which are poorly amenable to traditional biochemical approaches like immunoprecipitation and tandem affinity purification. Yet, these methods have so far not been applied extensively to invertebrate experimental models such as C. elegans given the long labeling times required for the original promiscuous biotin ligase variant BirA*. Here, we show that the recently developed variant TurboID successfully probes the interactomes of both stably associated (SPD-5) and dynamically localized (PLK-1) centrosomal components. We further develop an indirect proximity labeling method employing a GFP nanobody- TurboID fusion, which allows the identification of protein interactors in a tissue-specific manner in the context of the whole animal. Critically, this approach utilizes available endogenous GFP fusions, avoiding the need to generate multiple additional strains for each target protein and the potential complications associated with overexpressing the protein from transgenes. Using this method, we identify homologs of two highly conserved centriolar components, Cep97 and Bld10/Cep135, which are present in various somatic tissues of the worm. Surprisingly, neither protein is expressed in early embryos, likely explaining why these proteins have escaped attention until now. Our work expands the experimental repertoire for C. elegans and opens the door for further studies of tissue-specific variation in centrosome architecture.

Download Full-text

PathwayMatcher: proteoform-centric network construction enables fine-granularity multiomics pathway mapping

GigaScience ◽

10.1093/gigascience/giz088 ◽

2019 ◽

Vol 8 (8) ◽

Author(s):

Luis Francisco Hernández Sánchez ◽

Bram Burger ◽

Carlos Horro ◽

Antonio Fabregat ◽

Stefan Johansson ◽

...

Keyword(s):

Protein Interactions ◽

Posttranslational Modifications ◽

Knowledge Bases ◽

Protein Isoforms ◽

Biomedical Data ◽

Protein Protein Interactions ◽

Gene Sets ◽

Functional Knowledge ◽

Pathway Analyses ◽

Different Levels

Abstract Background Mapping biomedical data to functional knowledge is an essential task in bioinformatics and can be achieved by querying identifiers (e.g., gene sets) in pathway knowledge bases. However, the isoform and posttranslational modification states of proteins are lost when converting input and pathways into gene-centric lists. Findings Based on the Reactome knowledge base, we built a network of protein-protein interactions accounting for the documented isoform and modification statuses of proteins. We then implemented a command line application called PathwayMatcher (github.com/PathwayAnalysisPlatform/PathwayMatcher) to query this network. PathwayMatcher supports multiple types of omics data as input and outputs the possibly affected biochemical reactions, subnetworks, and pathways. Conclusions PathwayMatcher enables refining the network representation of pathways by including proteoforms defined as protein isoforms with posttranslational modifications. The specificity of pathway analyses is hence adapted to different levels of granularity, and it becomes possible to distinguish interactions between different forms of the same protein.

Download Full-text

mCSM-PPI2: predicting the effects of mutations on protein–protein interactions

Nucleic Acids Research ◽

10.1093/nar/gkz383 ◽

2019 ◽

Vol 47 (W1) ◽

pp. W338-W344 ◽

Cited By ~ 51

Author(s):

Carlos H M Rodrigues ◽

Yoochan Myung ◽

Douglas E V Pires ◽

David B Ascher

Keyword(s):

Protein Interactions ◽

Interaction Network ◽

Evolutionary Information ◽

Biological Processes ◽

Missense Mutations ◽

Protein Protein Interactions ◽

Protein Protein Interaction ◽

Residue Interaction ◽

User Friendly ◽

Network Metrics

AbstractProtein–protein Interactions are involved in most fundamental biological processes, with disease causing mutations enriched at their interfaces. Here we present mCSM-PPI2, a novel machine learning computational tool designed to more accurately predict the effects of missense mutations on protein–protein interaction binding affinity. mCSM-PPI2 uses graph-based structural signatures to model effects of variations on the inter-residue interaction network, evolutionary information, complex network metrics and energetic terms to generate an optimised predictor. We demonstrate that our method outperforms previous methods, ranking first among 26 others on CAPRI blind tests. mCSM-PPI2 is freely available as a user friendly webserver at http://biosig.unimelb.edu.au/mcsm_ppi2/.

Download Full-text

Systematic identification of signal integration by protein kinase A

Proceedings of the National Academy of Sciences ◽

10.1073/pnas.1409938112 ◽

2015 ◽

Vol 112 (14) ◽

pp. 4501-4506 ◽

Cited By ~ 37

Author(s):

Marie Filteau ◽

Guillaume Diss ◽

Francisco Torres-Quiroz ◽

Alexandre K. Dubé ◽

Andrea Schraffl ◽

...

Keyword(s):

Protein Kinase ◽

Protein Kinase A ◽

Protein Interactions ◽

Carbon Sources ◽

Biological Processes ◽

Protein Protein Interactions ◽

Cellular Processes ◽

Protein Complex Formation ◽

Growth And Aging ◽

Kinase A

Cellular processes and homeostasis control in eukaryotic cells is achieved by the action of regulatory proteins such as protein kinase A (PKA). Although the outbound signals from PKA directed to processes such as metabolism, growth, and aging have been well charted, what regulates this conserved regulator remains to be systematically identified to understand how it coordinates biological processes. Using a yeast PKA reporter assay, we identified genes that influence PKA activity by measuring protein–protein interactions between the regulatory and the two catalytic subunits of the PKA complex in 3,726 yeast genetic-deletion backgrounds grown on two carbon sources. Overall, nearly 500 genes were found to be connected directly or indirectly to PKA regulation, including 80 core regulators, denoting a wide diversity of signals regulating PKA, within and beyond the described upstream linear pathways. PKA regulators span multiple processes, including the antagonistic autophagy and methionine biosynthesis pathways. Our results converge toward mechanisms of PKA posttranslational regulation by lysine acetylation, which is conserved between yeast and humans and that, we show, regulates protein complex formation in mammals and carbohydrate storage and aging in yeast. Taken together, these results show that the extent of PKA input matches with its output, because this kinase receives information from upstream and downstream processes, and highlight how biological processes are interconnected and coordinated by PKA.

Download Full-text

Targeting protein–protein interactions by rational design: mimicry of protein surfaces

Journal of The Royal Society Interface ◽

10.1098/rsif.2006.0115 ◽

2006 ◽

Vol 3 (7) ◽

pp. 215-233 ◽

Cited By ~ 113

Author(s):

Steven Fletcher ◽

Andrew D Hamilton

Keyword(s):

Protein Interactions ◽

Active Site ◽

Rational Design ◽

Considerable Progress ◽

Biological Processes ◽

Protein Protein Interactions ◽

Protein Surfaces ◽

Enzyme Active Site ◽

Novel Therapeutics ◽

Interfacial Areas

Protein–protein interactions play key roles in a range of biological processes, and are therefore important targets for the design of novel therapeutics. Unlike in the design of enzyme active site inhibitors, the disruption of protein–protein interactions is far more challenging, due to such factors as the large interfacial areas involved and the relatively flat and featureless topologies of these surfaces. Nevertheless, in spite of such challenges, there has been considerable progress in recent years. In this review, we discuss this progress in the context of mimicry of protein surfaces: targeting protein–protein interactions by rational design.

Download Full-text

Computationally designed high specificity inhibitors delineate the roles of BCL2 family proteins in cancer

eLife ◽

10.7554/elife.20352 ◽

2016 ◽

Vol 5 ◽

Cited By ~ 40

Author(s):

Stephanie Berger ◽

Erik Procko ◽

Daciana Margineantu ◽

Erinna F Lee ◽

Betty W Shen ◽

...

Keyword(s):

Protein Interactions ◽

Human Cancer ◽

High Specificity ◽

Specific Protein ◽

Biological Processes ◽

Protein Protein Interactions ◽

Human Cancer Cell Lines ◽

Bcl2 Family ◽

Bcl2 Protein

Many cancers overexpress one or more of the six human pro-survival BCL2 family proteins to evade apoptosis. To determine which BCL2 protein or proteins block apoptosis in different cancers, we computationally designed three-helix bundle protein inhibitors specific for each BCL2 pro-survival protein. Following in vitro optimization, each inhibitor binds its target with high picomolar to low nanomolar affinity and at least 300-fold specificity. Expression of the designed inhibitors in human cancer cell lines revealed unique dependencies on BCL2 proteins for survival which could not be inferred from other BCL2 profiling methods. Our results show that designed inhibitors can be generated for each member of a closely-knit protein family to probe the importance of specific protein-protein interactions in complex biological processes.

Download Full-text

DISCOVERING PROTEIN–PROTEIN INTERACTIONS

Journal of Bioinformatics and Computational Biology ◽

10.1142/s0219720004000600 ◽

2004 ◽

Vol 01 (04) ◽

pp. 711-741 ◽

Cited By ~ 9

Author(s):

SEE-KIONG NG ◽

SOON-HENG TAN

Keyword(s):

Protein Interactions ◽

Experimental Methods ◽

Protein Interaction Networks ◽

Interaction Networks ◽

Biological Processes ◽

Protein Protein Interactions ◽

Comprehensive Description ◽

New Genes ◽

Genomics And Proteomics ◽

Living Organisms

The ongoing genomics and proteomics efforts have helped identify many new genes and proteins in living organisms. However, simply knowing the existence of genes and proteins does not tell us much about the biological processes in which they participate. Many major biological processes are controlled by protein interaction networks. A comprehensive description of protein–protein interactions is therefore necessary to understand the genetic program of life. In this tutorial, we provide an overview of the various current high-throughput methods for discovering protein–protein interactions, covering both the conventional experimental methods and new computational approaches.

Download Full-text

PathwayMatcher: proteoform-centric network construction enables fine-granularity multi-omics pathway mapping

10.1101/375097 ◽

2018 ◽

Author(s):

Luis Francisco Hernández Sánchez ◽

Bram Burger ◽

Carlos Horro ◽

Antonio Fabregat ◽

Stefan Johansson ◽

...

Keyword(s):

Protein Interactions ◽

Biomedical Data ◽

Protein Protein Interactions ◽

Post Translational Modification ◽

Post Translational Modifications ◽

Network Construction ◽

Gene Sets ◽

Functional Knowledge ◽

Pathway Analyses ◽

Different Levels

AbstractBackgroundMapping biomedical data to functional knowledge is an essential task in bioinformatics and can be achieved by querying identifiers, e.g. gene sets, in pathway knowledgebases. However, the isoform and post-translational modification states of proteins are lost when converting input and pathways into gene-centric lists.FindingsBased on the Reactome knowledgebase, we built a network of protein-protein interactions accounting for the documented isoform and modification statuses of proteins. We then implemented a command line application called PathwayMatcher (github.com/PathwayAnalysisPlatform/PathwayMatcher) to query this network. PathwayMatcher supports multiple types of omics data as input, and outputs the possibly affected biochemical reactions, subnetworks, and pathways.ConclusionsPathwayMatcher enables refining the network-representation of pathways by including isoform and post-translational modifications. The specificity of pathway analyses is hence adapted to different levels of granularity and it becomes possible to distinguish interactions between different forms of the same protein.

Download Full-text