Representative Proteomes: A Stable, Scalable and Unbiased Proteome Set for Sequence Analysis and Functional Annotation

By Chuming Chen, D. Natale, Robert D. Finn, Hongzhan Huang, Jian Zhang, Cathy H. Wu, R. Mazumder. 2011

113 citations recorded by the source corpus.

The accelerating growth in the number of protein sequences taxes both the computational and manual resources needed to analyze them. One approach to dealing with this problem is to minimize the number of proteins subjected to such analysis in a way that minimizes loss of information. To this end we have developed a set of Representative Proteomes (RPs), each selected from a Representative Proteome Group (RPG) containing similar proteomes calculated based on co-membership in UniRef50 clusters. A Representative Proteome is the proteome that can best represent all the proteomes in its group in terms of the majority of the sequence space and information. RPs at 75%, 55%, 35% and 15% co-membership threshold (CMT) are provided to allow users to decrease or increase the granularity of the sequence space based on their requirements. We find that a CMT of 55% (RP55) most closely follows standard taxonomic classifications. Further analysis of this set reveals that sequence space is reduced by more than 80% relative to UniProtKB, while retaining both sequence diversity (over 95% of InterPro domains) and annotation information (93% of experimentally characterized proteins). All sets can be browsed and are available for sequence similarity searches and download at http://www.proteininformationresource.org/rps, while the set of 637 RPs determined using a 55% CMT are also available for text searches. Potential applications include sequence similarity searches, protein classification and targeted protein annotation and characterization.

Subfields: Molecular Biology, Spectroscopy

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

In Spectroscopy

Citations in the corpus

UniProt: the universal protein knowledgebase (2016, methodology): “These reference proteomes are selected via consulta-tion with the research community or computationally determined from proteome clusters (7) using an algorithm that considers the best overall annotation score, an automatically calculated score which provides a heuristic measure of the annotation…”
UniRef clusters: a comprehensive and scalable alternative for improving sequence similarity searches (2014, methodology): “…sequence search sensitivity with transitive alignments (Malde and Furmanek, 2013), developing representative proteomes and proteome clusters (Chen et al., 2011), predicting the functional effects of disease variants (Capriotti and Altman, 2011a, b; Sim et al., 2012), performing functional…”
UniProt: a worldwide hub of protein knowledge (2018, methodology): “For the remaining proteomes we provide a Reference Pro-teome set ( ∼ 9% of total proteomes) selected by the research community and supplemented with selected proteomes from a computational clustering (6) to provide the best annotated proteome in their cluster.”
The Pfam protein families database (2011, background): “(B) Number of PIR representative clusters of genomes (23) in DUF families.”
The Pfam protein families database: towards a more sustainable future (2015, methodology): “In the last NAR article describing the Pfam database (7), the use of representative proteomes (RP) (8) to provide non-redundant versions of the full alignments was described.”
UniProt: a hub for protein information (2014, methodology): “UniProt reference proteomes are derived via consultation with the research community or computationally determined from proteome clusters (5) where the reference proteome is selected from the cluster by an algorithm that considers the best overall annotation score.”
HMMER web server: 2018 update (2018, methodology): “The sequence databases are largely based on UniProtKB (11), with the option of searching either this database in its entirety, or one of the various subsets: UniProtKB Reference Proteomes, Representative Proteomes (12) or the curated sequences of UniProtKB/SwissProt.”
Connecting the Sequence-Space of Bacterial Signaling Proteins to Phenotypes Using Coevolutionary Landscapes (2016, methodology): “We searched the Representative Proteomes (RP55) (51) database for HK proteins using the Jackhmmer algorithm on the HMMER web server (52) with default parameters.”
Update on activities at the Universal Protein Resource (UniProt) in 2013 (2012, background): “In response, the UniProt Consortium is developing a concept to provide a set of sequences from selected species based on UniProtKB manually reviewed entries, the Reference Proteomes and the Representative proteomes (3,4).”
Advances and Applications in the Quest for Orthologs (2019, background): “This was done to broaden the species coverage, by using the Proteome Priority Score (Chen et al. 2011) to identify the best species candidate for a given phylum.”
Pfam: The protein families database in 2021 (2020, background): “Similarly, the match data for representative proteomes (5) are provided in the same way.”
PANTHER version 16: a revised family classification, tree-based classification tool, enhancer regions and extensive API (2020, methodology): “PANTHER families are constructed each year, using the latest protein sequences available from UniProt Reference Proteomes (15,18).”
HMMER web server: 2015 update (2015, methodology): “While searches remain fast, returning in a matter of a few seconds, subsets of UniProtKB have been included to provide either the highest quality (UniProtKB/Swiss-Prot) annotations, representative sets to provide good coverage of sequence space while reducing the number of potential matches (UniProt”
Minimizing proteome redundancy in the UniProt Knowledgebase (2016, methodology): “Representative Proteome Groups defined by the co-membership in UniRef clusters (9) were used to aid the selection of our Reference proteomes.”
Reorganizing the protein space at the Universal Protein Resource (UniProt) (2011, background): “A RP is the proteome that can best represent all the proteomes in its group in terms of the majority of the sequence space and annotation (7).”
Pfam: the protein families database (2013, methodology): “To provide more useable samples of the sequence diversity within a family, we now calculate model-matches for four additional sequence sets, based on ‘Representative Proteomes’ (RPs) (13).”