Analysis of the Evolving Proteomes: Predictions of the Number of Protein Domains in Nature and the Number of Genes in Eukaryotic Organisms

Abstract

Motivation: Obtaining accurate estimates of the numbers of protein-coding genes and protein domains in a proteome, and the number of protein domains in nature is a daunting challenge. Computational analysis of the protein domain sets in the proteomes of many species allows us to estimate these numbers and to find their evolution relationships.Results: We have analyzed the distributions of the number of occurrences of protein domains in sample proteomes of the 70 fully sequenced genome organisms of three major kingdoms of life: Archaea, Bacteria and Eukaryota. We found that a large fraction of the identified distinct protein domains (i.e., unique domains and homologous domain families) in these 70 proteomes (1051 (23%) out of 4493) are found in at least one organism in each of these kingdoms of life and that 43 (1%) of these domains are common to all the 70 organisms. All the observed domain occurrence frequency distributions for these 70 proteomes are well fitted by a family of Pareto-like functions, associated with the steady state distributions of a linear Markov random process. We present explicit * Corresponding author. 381 J. Biol. Syst. 2002.10:381-407. Downloaded from www.worldscientific.com by VIRGINIA COMMONWEALTH UNIV on 03/13/15. For personal use only. 382 Kuznetsov, Pickalov, Senko & Knott formulas that accurately predict the number of distinct protein domains and the number of protein-coding genes for a given organism as functions of the number of non-redundant domain-to-protein links in the proteomes. These functions allows us to predict that there are 42,740, 27,900, and 21,200 protein-coding genes/open reading frames in the human,A. thaliana, and mouse genomes, respectively. We also estimate that there are 5271, 2955, and 4915 distinct protein domains in the human, A. thaliana, and mouse proteomes, respectively, and about 5500 distinct protein domains in the entire "proteome world".

Projection of gene-protein networks to the functional space of the proteome and its application to analysis of organism complexity

“…This distribution suggests an appearance of hub nodes in the corresponding functional networks of proteins. Due to the non-zero value parameter b , the network is a scale dependent network [ 8 , 18 ]. This indicates an additional mechanistic source of functional complexity of interconnections of splice variants and protein functions in human.…”

“…Note, similar statistics could emerge in the random graphs when a nodal degree distribution in the graph follows asymptotically (at large enough m ) GDP function with constant exponent ( k+1 ), where k+1 ≈3 and b =0 [ 17 ]. The distribution function of the events could be described as a stochastic process, which has no or very minor influence on restrictions and/or evolution-driven connections [ 7 - 9 , 18 , 25 - 27 ]. Thus, by a scale-free network model, a new biological function could be acquired or lost by a TU almost arbitrarily with or without selection pressure.…”

Evolution of Protein Domain Architectures

“…They found that a majority of domain families are common to all three kingdoms of life and thus likely to be ancient. Kuznetsov et al [43] performed a similar analysis using InterPro domains and found that only about one fourth of all such domains were present in all three kingdoms, but a majority was present in more than one of them. Lateral gene transfer or annotation errors can cause a domain family to be found in one or a few species in a kingdom without actually belonging to that kingdom.…”

“…Later work ( [39,43], see also [44]) demonstrated that proteome-wide domain occurrence data fit the general GPD better than the power law but that it also asymptotically fits a power law as X ) i. The deviation from strict power law behavior depends on proteome size in a kingdom-dependent manner [43]. Regardless, it is mostly appropriate to treat the domain family size distribution as approximately (and asymptotically) power law-like, and later studies typically assume this.…”

“…Later, more complex birth-death [43] and birth-death-andinnovation (BDIM) [29,34,39,46] models were introduced to explain the observed distributions, and from investigating which model parameter ranges allow this fit, the authors were able to draw several far-ranging conclusions. First, the asymptotic power law behavior requires that the rates of domain gain and loss are asymptotically equal.…”