Mapping high-quality binary protein-protein interactomes
Human Reference Interactome Map - HuRI Project
The study of human cells as a system requires near-complete catalogs of all cellular parts, e.g. the genes, proteins, transcripts, organelles, etc. but equally importantly we also need comprehensive maps about how these cellular components interact with each other to mediate cellular functions. Proteins are considered the master regulators of cellular functions by interacting with each other and with many other biological (macro)molecules, yet, available protein-protein interaction (PPI) datasets only cover a small fraction of the estimated human protein interactome.
One of the long-term goals at the Center for Cancer Systems Biology (CCSB) is to generate the first reference map of
the binary human protein interactome network. To reach this target, we are identifying binary PPIs by systematically interrogating all pairwise combinations of predicted human protein-coding genes using proteome-scale technologies. Our approach to map high-quality PPIs is based on using yeast two-hybrid (Y2H) as the primary screening method followed by validation of subsets of PPIs in multiple orthogonal assays for binary PPI detection.
The mapping of human PPIs at CCSB has occurred in multiple stages where each effort is defined by the technology and set of human Open Reading Frames (ORFs) available at that time for the screening. Two efforts encompassing a search space of ~7,000 and 13,000 human genes, respectively, have already been published identifying in total ~16,000 binary PPIs (Rual, et al. Nature 2015 Rolland, et al. Cell 2014). In an ongoing effort, ORFs representing ~18,000 human genes are being screened against each other for PPI identification by employing multiple Y2H assay variants to increase the overall number of PPIs identified (Luck, et al. Nature 2020). All human PPIs identified in systematic screens at CCSB, currently ~50,000 PPIs, are available for search and download at http://interactome-atlas.org/.
Please contact Michael Calderwood with your questions and comments.
The study of human cells as a system requires near-complete catalogs of all cellular parts, e.g. the genes, proteins, transcripts, organelles, etc. but equally importantly we also need comprehensive maps about how these cellular components interact with each other to mediate cellular functions. Proteins are considered the master regulators of cellular functions by interacting with each other and with many other biological (macro)molecules, yet, available protein-protein interaction (PPI) datasets only cover a small fraction of the estimated human protein interactome.
One of the long-term goals at the Center for Cancer Systems Biology (CCSB) is to generate the first reference map of
the binary human protein interactome network. To reach this target, we are identifying binary PPIs by systematically interrogating all pairwise combinations of predicted human protein-coding genes using proteome-scale technologies. Our approach to map high-quality PPIs is based on using yeast two-hybrid (Y2H) as the primary screening method followed by validation of subsets of PPIs in multiple orthogonal assays for binary PPI detection.
The mapping of human PPIs at CCSB has occurred in multiple stages where each effort is defined by the technology and set of human Open Reading Frames (ORFs) available at that time for the screening. Two efforts encompassing a search space of ~7,000 and 13,000 human genes, respectively, have already been published identifying in total ~16,000 binary PPIs (Rual, et al. Nature 2015 Rolland, et al. Cell 2014). In an ongoing effort, ORFs representing ~18,000 human genes are being screened against each other for PPI identification by employing multiple Y2H assay variants to increase the overall number of PPIs identified (Luck, et al. Nature 2020). All human PPIs identified in systematic screens at CCSB, currently ~50,000 PPIs, are available for search and download at http://interactome-atlas.org/.
Please contact Michael Calderwood with your questions and comments.
Yeast Reference Interactome Map - YeRI Project
The model organism, Saccharomyces cerevisiae, is one of the best-characterized species at the level of systems biology. Many large-scale datasets, examining binary protein-protein interactions, genetic interactions, and protein co-complex associations, among others, have been released for budding yeast. This is in contrast to system-level datasets for human, of which few have been released. With a small and easily tractable genome, yeast serves as a powerful model system for genome-wide and proteome-wide studies. Moreover, many biological pathways that are of interest in human cells are conserved in yeast.
Our recent mapping effort seeks to expand our current knowledge of the protein-protein interactome network in yeast. Despite several systematic mapping efforts published within the last few decades, only a fraction of predicted binary PPIs have been identified. Our efforts to expand the PPI network map rely on several key parameters. First, we have increased the search space, testing all of the ~6,500 annotated yeast genes. Additionally, we have screened yeast proto-genes, which are open reading frames that are annotated as non-coding, but have evidence of transcription and translation. Identified computationally, proto-genes are transitory intermediates in the mechanism of de novo gene birth (Figure 1), and have never been examined for the ability to form PPIs. This mapping effort will therefore be the most comprehensive effort in terms of search space and the number of proteins queried.
The model organism, Saccharomyces cerevisiae, is one of the best-characterized species at the level of systems biology. Many large-scale datasets, examining binary protein-protein interactions, genetic interactions, and protein co-complex associations, among others, have been released for budding yeast. This is in contrast to system-level datasets for human, of which few have been released. With a small and easily tractable genome, yeast serves as a powerful model system for genome-wide and proteome-wide studies. Moreover, many biological pathways that are of interest in human cells are conserved in yeast.
Our recent mapping effort seeks to expand our current knowledge of the protein-protein interactome network in yeast. Despite several systematic mapping efforts published within the last few decades, only a fraction of predicted binary PPIs have been identified. Our efforts to expand the PPI network map rely on several key parameters. First, we have increased the search space, testing all of the ~6,500 annotated yeast genes. Additionally, we have screened yeast proto-genes, which are open reading frames that are annotated as non-coding, but have evidence of transcription and translation. Identified computationally, proto-genes are transitory intermediates in the mechanism of de novo gene birth (Figure 1), and have never been examined for the ability to form PPIs. This mapping effort will therefore be the most comprehensive effort in terms of search space and the number of proteins queried.
|
|
|
|
Moreover, all of the known binary PPIs were identified by Y2H using similar low-copy-number screening vectors. We have shown that using alternate screening vectors, particularly high-copy-number plasmids, allows for the identification of previously undetected pairs of known interacting proteins (Figure 2) (positive reference set (PRS) consisting of interacting protein pairs reported in the literature and random reference set (RRS) consisting of random protein pairs not reported in the literature), presumably by modulating the expression levels of the protein pairs that are being tested. We are currently examining our data to determine how this mapping effort can enrich our current knowledge of yeast biology, and how such information can help us understanding human disease mechanisms.
Please contact Luke Lambourne with your questions and comments.
Please contact Luke Lambourne with your questions and comments.
Drosophila Interactome Map - FlyBi Project
The Drosophila melanogaster genome is one of the best-annotated multi-cellular eukaryotic genomes and yet our knowledge of protein-protein interactions (PPIs), protein complexes and networks in D. melanogaster proteomics is still limited. Drosophila is an important model for disease-focused and basic biological studies and can be used, for example, for the study of conserved gene functions related to cancers, neurodegenerative diseases, diabetes. D. melanogaster also serves as an important model for the study of insect pests and disease vectors such as mosquitos.
Together with the Berkeley Drosophila Genome Project (BDGP) and the Drosophila RNAi Screening Center (DRSC) we propose a state-of-the-art, high-throughput, quality-controlled binary protein interaction analysis with ~10,000 D. melanogaster open reading frames (ORFs), representing about two thirds of the proteome, to generate a high-confidence binary protein-protein interaction network.
Using our established binary interaction mapping pipeline, we are screening 10,000 x 10,000 ORFs pairs for interactions with one another, totaling approximately ~100 million pairs. Primary screening and pairwise verification assays are performed using the Y2H platform, followed by validation of the dataset using orthogonal orthogonal binary interaction assays, such as MAPPIT and GPCA.
For more information about the FlyBi project, including information regarding distribution of FlyBi clones from our repositories, please see http://flybi.hms.harvard.edu/
Nature Communications, Tang & Spirohn et al.
Please contact Kerstin Spirohn-Fitzgerald with your questions and comments.
The Drosophila melanogaster genome is one of the best-annotated multi-cellular eukaryotic genomes and yet our knowledge of protein-protein interactions (PPIs), protein complexes and networks in D. melanogaster proteomics is still limited. Drosophila is an important model for disease-focused and basic biological studies and can be used, for example, for the study of conserved gene functions related to cancers, neurodegenerative diseases, diabetes. D. melanogaster also serves as an important model for the study of insect pests and disease vectors such as mosquitos.
Together with the Berkeley Drosophila Genome Project (BDGP) and the Drosophila RNAi Screening Center (DRSC) we propose a state-of-the-art, high-throughput, quality-controlled binary protein interaction analysis with ~10,000 D. melanogaster open reading frames (ORFs), representing about two thirds of the proteome, to generate a high-confidence binary protein-protein interaction network.
Using our established binary interaction mapping pipeline, we are screening 10,000 x 10,000 ORFs pairs for interactions with one another, totaling approximately ~100 million pairs. Primary screening and pairwise verification assays are performed using the Y2H platform, followed by validation of the dataset using orthogonal orthogonal binary interaction assays, such as MAPPIT and GPCA.
For more information about the FlyBi project, including information regarding distribution of FlyBi clones from our repositories, please see http://flybi.hms.harvard.edu/
Nature Communications, Tang & Spirohn et al.
Please contact Kerstin Spirohn-Fitzgerald with your questions and comments.