The Pfam protein families database
By Robert D. Finn, Jaina Mistry, John G. Tate, Penny C. Coggill, Andreas Heger, Joanne E. Pollington, O. Gavin, P. Gunasekaran, Goran Ceric, Kristoffer Forslund and 4 more. 2007
15,542 citations recorded by the source corpus.
Pfam is a widely used database of protein families and domains. This article describes a set of major updates that we have implemented in the latest release (version 24.0). The most important change is that we now use HMMER3, the latest version of the popular profile hidden Markov model package. This software is ∼100 times faster than HMMER2 and is more sensitive due to the routine use of the forward algorithm. The move to HMMER3 has necessitated numerous changes to Pfam that are described in detail. Pfam release 24.0 contains 11 912 families, of which a large number have been significantly updated during the past two years. Pfam is available via servers in the UK (http://pfam.sanger.ac.uk/), the USA (http://pfam.janelia.org/) and Sweden (http://pfam.sbc.su.se/).
Subfields: Molecular Biology, Spectroscopy, Computational Biology & Bioinformatics
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
- Yang Zhang
- Jinbo Xu
- David T. Jones
- Andrej Sali
- Johannes Söding
- Burkhard Rost
- Sheng Wang
- Torsten Schwede
In Spectroscopy
- Burkhard Rost
- Sean R. Eddy
- Johannes Söding
- Jinbo Xu
- David T. Jones
- Sheng Wang
- Matthias Mann
- Jüergen Cox
In Computational Biology & Bioinformatics
- Yang Zhang
- Jinbo Xu
- David T. Jones
- Sheng Wang
- Johannes Söding
- Burkhard Rost
- Arne Elofsson
- Martin Weigt
Citations in the corpus
- Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences (2019, methodology): “We estimate differences between the aligned and unaligned distributions across 128 Pfam families using the area under the ROC curve (AUC) as a metric of discriminative power between aligned and unaligned pairs.”
- ECOD: An Evolutionary Classification of Protein Domains (2014, background): “However, not all protein domains with known structure can be attributed to the current version of Pfam by sequence similarity.”
- Hidden Markov model speed heuristic and iterative HMM search procedure (2010, background): “Current datasets of homologous proteins primarily utilize either protein structure or sequence conservation to infer homology [8-10].”
- Learning protein sequence embeddings using information from structure (2019, methodology): “The language model is pretrained on the raw protein sequences in the protein families database (Pfam) [35] to predict the amino acid at each position of each protein given the previous amino acids and the following amino acids (see Appendix section A.”
- CCMpred—fast and precise prediction of protein residue–residue contacts from correlated mutations (2014, background): “We estimate that roughly a third of all families in the Pfam database (Finn et al., 2014) can already be predicted using a reasonable sequence coverage threshold of N/L ≥ 3.”
- Protein 3D Structure Computed from Evolutionary Sequence Variation (2011, methodology): “families), have a 3D structure for any of their members [1].”
- Direct-coupling analysis of residue coevolution captures native contacts across many protein families (2011, methodology): “…multiple sequence alignment (MSA) of a large number of sequences of a given protein domain, extracted using Pfam’s Hidden Markov Models (HMMs) (21, 22), the basic quantities in this context are the frequency count The raw statistical correlation obtained above suffers from a sampling bias,…”
- Learning generative models for protein fold families (2011, methodology): “The well-known database PFAM [4], for example, contains more than 11,000 profile Hidden Markov Models (HMM) [6] learned from MSAs.”
- Identification of direct residue contacts in protein–protein interaction by message passing (2009, positive): “domains,asdefinedbyhiddenMarkovmodels (HMMs) inthePfamdatabase(40),”
- Protein structure prediction from sequence variation (2012, background): “A reasonable extrapolation predicts that within a few years most of the current 15,000 protein families (as defined by PFAM-A 55 ) will have sufficiently many known sequences to yield a robust evolutionary coupling signal ( Fig.”
- Accelerated Profile HMM Searches (2011, methodology): “Example sequence alignments and profile HMMs were sampled from Seed alignments and profiles in Pfam 24 [11].”
- The Pfam protein families database (2011, background): “The list is the same as that reported in 2010 (4), except that Bacillus subtilis has also been included.”
- Continuous Distributed Representation of Biological Sequences for Deep Proteomics and Genomics (2015, neutral): “Protein Family Database (Pfam) is a widely used source for protein families [15].”
- The Pfam protein families database: towards a more sustainable future (2015, methodology): “The profile HMM is searched against sequence databases, with all matches scoring greater than or equal to a curated threshold (the gathering threshold, previously described in detail (2)) being considered as true members.”