Lukasz Jaroszewski; Zhanwen Li; S. Sri Krishna; Constantina Bakolitsa; John Wooley; Ashley M. Deacon; Ian A. Wilson; Adam Godzik

“ Thus, the most likely explanation of this phenomenon is that, with an increasing number of known protein structures, we are observing a saturation of the available fold space that is first seen at the level of micro-domains that represent shorter, usually compact, structures that become component pieces of different folds. Therefore, we would contend that most similarities in this group are examples of “constrained” evolution, where similar solutions are found independently because of the limited number of choices available. ”

“ The growth of the protein universe can then be investigated on the level of individual proteins, protein families, or clans/superfamilies, and we can expect qualitatively different answers on each level. ”

“ Instead of focusing on individual proteins, US structural genomics and, specifically, the four large-scale production centers of the PSI have focused their attention on substantially increasing structural coverage of protein space. DUF families have then become natural targets as such families cover a significant fraction of the unexplored protein universe. In contrast, “classical” structural biology efforts are mainly focused on well-characterized systems, leaving the majority of protein families outside of their sphere of interest, including, by default, almost all DUF families. ”

“ The PSI has determined the first structures for more than 600 protein families that were selected by novelty of their sequences and lack of any structural information. These efforts have made it abundantly clear that the vast majority of structurally uncharacterized protein space consists of families that will be eventually classified into already known folds. The most recent results from the PSI suggest that this trend is accelerating and the percentage of new folds in structures representing new families solved in 2008 is substantially lower than the historical trend. ”

“ In the group of “putative analogs”—the 62 DUF families with known folds, but no other detectable homology to other proteins—hypotheses about function can presently be based only on the structural information. In the first approximation, the possibility of inferring function from the protein fold depends on the numbers and functional diversity of proteins that have already been observed to adopt this fold. Some protein folds, although highly populated, are strongly associated with a specific function or functional category, while other folds contain proteins of diverse functions. ”

“ The surprisingly small, although certainly not insignificant, percentage of new folds among the DUF families is a major result of this study. It is also interesting to note that this modest percentage of new folds among DUF families has been steadily decreasing over time—over 50% of the DUFs analyzed in this study were solved in the last 2 y, but they contain only 20% of the new folds found in the 248 DUFs. We would certainly still expect to find new folds as we continue to explore protein sequence space, as our knowledge of the protein structure universe is far from complete. ”

“ Evidence of saturation of protein fold space as a function of time.With growing number of folds, the percentage of folds with partial structural similarity to other folds is increasing, and hence, the number of truly new folds being discovered is rapidly decreasing. ”

“ Structural similarity validates (or invalidates) distant homology predictions simply by verifying whether it had linked the DUF family to the correct (or incorrect) fold. As a result, homology predictions validated by solved structures make hypotheses about function more reliable and more precise.
Our statistics can be also used to revisit the age old question as to the number of protein folds, which have been estimated by several different approaches to be between 1,000 [38] and 10,000 [30] – [31] . ”

“ The sequences of several millions of proteins are currently known and this number is growing ever more rapidly as a result of the relentless efficiency of genomic and metagenomic sequencing projects. Around 30%–40% of these gene products are classified as so-called “hypothetical proteins.” This term is somewhat of a misnomer, but is the accepted way of indicating that no information is available about them other than the translated nucleotide sequence. It is interesting to note that this group of proteins persists despite years of annotation efforts in the genome sequencing projects. ”

“ These families are now being systematically added to new releases of the PFAM database (see statistics and discussion about DUF families at http://xfam.wordpress.com/) . The PSI centers have also solved structural representatives of over 300 protein families that have only limited functional information. However, for clarity and for ease of comparison to PFAM and Structural Classification of Proteins (SCOP) annotations, we have focused our analyses here on the group of 248 protein families identified and classified by PFAM as DUFs and already classified by SCOP for their structural novelty. ”

“ Abstract Determination of first protein structures, from hundreds of families of unknown function, have shown that divergence, rather than novelty, is the dominant force that shapes the evolution of the protein universe.==Abstract== The genome projects have unearthed an enormous diversity of genes of unknown function that are still awaiting biological and biochemical characterization. These genes, as most others, can be grouped into families based on sequence similarity. ”

“ We note that, by definition, the set of partial similarities is limited to pairs with more than 50 equivalent residues and Cα RMSD below 3 Å. Divergence versus Novelty
Each of the 248 DUF structures solved by PSI was analyzed by structure comparison and remote homology recognition tools (see Materials and Methods) . In contrast to their designation, 25% of these DUF families can be linked to other protein families using a sensitive, profile-profile alignment algorithm [24] or other similar tools at a significance level where we expect <5% false predictions. ”

“ While rigorous proof of homology is often difficult, if not impossible, usually the combination of several arguments enables us to arrive at a satisfactory answer. Using this approach, we can propose that, for most of the cases investigated here, the similarities of these DUF families to known folds and families represent actual, albeit often very distant, homologies. ”

“ Such a realization correlates well with our understanding of molecular evolution, which often obliterates easily recognizable sequence similarity among genes that diverged a long time ago, but leaves behind traces of statistically significant patterns of conserved residues that are apparent only when multiple, related sequences are aligned. ”

“ Folds were added in historical order in groups of 100 and the percentage of folds with partial similarity to any previously solved fold was calculated for each group. All cases in which FATCAT algorithm found at least 50 equivalent residues superimposed with Cα RMSD <3 Å were regarded as putative cases of “significant partial similarity” and were subject to visual verification. ”

“ For instance, in PFAM [23] such groups are called “clans,” whereas “superfamily” is often used in other resources. We can expect that further development of even more sensitive algorithms for recognition of distant homologs would expand the list of clans or equivalent groupings in other classification systems. ”

“ Distribution of DUF structures with regard to structural similarity and homology to previously known structures. The main pie chart shows overall percentages of DUF families with new folds, new folds partially similar to previously known folds, putative analogs, putative homologs, and recognizable homologs. The inset pie charts show the percentage of DUF families with proposed hypothesis about function in each of these six categories. (B) Impact of solved structures on hypotheses about function proposed for DUF families. ”

“ In addition to the trend described above, analyses of genuinely novel structures that represent new folds show that a significant and growing percentage contains complex structural elements or substructures that are present in other folds. ”

“ Notwithstanding, these DUF families still constitute the richest source for discovery of the remaining protein folds and topologies.==Abstract== Author SummaryMore than 40% of known proteins lack any annotation within public databases and are usually referred to as hypothetical proteins despite most of them being real and many being evolutionarily conserved and thus expected to play important biological roles. ”

“ S4.17570143↑ Redfern O. C, Dessailly B, Orengo C. A. Exploring the structure and function paradigm. Curr Opin Struct Biol. 2008;18:394–402.18554899↑ Chothia C. Proteins. one thousand families for the molecular biologist. Nature. 1992;357:543–544.1608464↑ Kolodny R, Petrey D, Honig B. Protein structure comparison: implications for the nature of ‘fold space’, and structure and function prediction. Curr Opin Struct Biol. 2006;16:393–398.16678402↑ Honig B. Protein structure space is much more than the sum of its folds. ”

“ Comp Funct Genomics. 2001;2:14–18.18628897↑ Galperin M. Y, Koonin E. V. ‘Conserved hypothetical’ proteins: prioritization of targets for experimental study. Nucleic Acids Res. 2004;32:5452–5463.15479782↑ Pawlowski K. Uncharacterized/hypothetical proteins in biomedical ‘omics’ experiments: is novelty being swept under the carpet? Brief Funct Genomic Proteomic. 2008;7:283–290.18641417↑ Tatusov R. L, Koonin E. V, Lipman D. J. A genomic perspective on protein families. Science. 1997;278:631–637.9381173↑ Meinel T, Krause A, Luz H, Vingron M, Staub E. The SYSTERS Protein Family Database in 2005. ”