This week we are highlighting the three finest examples of proteomics data made public in 2013. As we have been doing since 2010, we are naming the best data in three categories. These ratings do not take into account the associated publication: only the data itself was considered in these awards. Any of these data sets would be ideal for use as standards in the development of computational biology methods associated with proteomics.

This data set consisted of 87 results, corresponding to either large LC/MS/MS runs of phosphopeptide enriched samples or small sets of spectra from individually excised 2D gel electrophoresis bands. The data files were made available through ProteomeXchange, PXD000340 (LC/MS/MS) and PXD000470 (2D). It has been published by Lv DW, Subburaj S, Cao M, Yan X, Li X, Appels R, Sun DF, Ma W and Yan YM, Mol Cell Proteomics. 2013 Dec 11 (PubMed).

This study highlights the utility of the model species Brachypodium distachyon (purple false brome), which has the potential to be used as a model for economically important temperate grasses, e.g. Triticum spp., in the same way that A. thaliana is used as a model for the economically important Brassicaceae species. Its genome and proteome have been sequenced and its relatively simple genetics (compared to the hexaploid wheat) make it a much more tractible platform for research. The data presented here were produced by well done experiments. It is also one of few publicly available studies that uses MALDI TOF/TOF to identify excised 2D gel bands. This type of data has a very different character than that generated by the more commonly used shotgun proteomics methods, making it interesting from an bioinformatics point-of-view. Rather than trying to unravel a complex mixture of peptides from all of the proteins in a sample, this type of data has a small number of spectra (typically ≤ 10) that correspond to a single protein species (splice variant/post-translationally modified sequence). Anyone interested in emphasizing the consequences of individual identifications should look carefully at this data, which may have considerable utility for the construction of tutorial material.

This data set consisted of 9 results, corresponding to LC/MS/MS analyses of samples enriched in phosphopeptides using IMAC purification. The data files were made available through ProteomeXchange, PXD000375. It has been published by Kao L, Wang YT, Chen YC, Tseng SF, Jhang JC, Chen YJ and Teng SC, Mol Cell Proteomics. 2013 Dec 7 (PubMed).

The results of this study demonstrate that IMAC sample preparation methods can be quite reproducible at both the protein and peptide levels. The biological question being examined was the converse of many studies that involve IMAC: rather than looking for kinase-induced protein phosphorylation, the study was designed to find phosphorylase-induced protein dephosphorylation events. The intepretation of the study was made more complicated by apparent HPLC timing problems in setting the start of mass spectrum data acquistion and the termination of the previous gradient.

This data set consisted of 80 results, generated from LC/MS/MS analyses of SDS-PAGE gel bands. The data files were made available through ProteomeXchange, PXD000067. It has been published by Bai B, Hales CM, Chen PC, Gozal Y, Dammer EB, Fritz JJ, Wang X, Xia Q, Duong DM, Street C, Cantero G, Cheng D, Jones DR, Wu Z, Li Y, Diner I, Heilman CJ, Rees HD, Wu H, Lin L, Szulwach KE, Gearing M, Mufson EJ, Bennett DA, Montine TJ, Seyfried NT, Wingo TS, Sun YE, Jin P, Hanfelt J, Willcock DM, Levey A, Lah JJ and Peng J, Proc Natl Acad Sci U S A. 2013 Oct 8;110(41):16562-7 (PubMed).

This study attempted to characterize the proteins present in the insoluble aggregates obtained from the brains of Alzheimer's Disease patients. The results demonstrated that a surprisingly large number of proteins were detectable in these aggregates, including many cytoskeletal, mitochondrial and ribonuclear proteins. The results were of sufficient quality that they can serve as a guide to the proteins expected in this type of preparation.

This data set consisted of 30 results, generated from affinity pull-down experiments. The data files were made available through ProteomeXchange, PXD000461. It has been published by Fujita T, Asano Y, Ohtsuka J, Takada Y, Saito K, Ohki R and Fujii H Sci Rep. 2013 Nov 8;3:3171 (PubMed).

This paper reports a method for obtaining proteins specifically enriched in those physically associated with a particular chromosomal feature, namely the telomere. The results suggests that the method works as designed: the resulting lists of proteins show significant enrichment in chromatin associated proteins, nuclear ribonucleoproteins as well as a number of relatively rarely observed proteins, such as Rhox5:p and Runx1t1:p. Some of the chromatograms show an odd dropout of signal in the latter half of the run. This puzzling feature does not diminish the potential utility of the method being demonstrated: it should encourage other groups to reproduce and improve upon the results reported.

This data set consisted of 84 results, generated from LC/MS/MS experiments. The data files were made available through ProteomeXchange, PXD000403. It has been published by Hartmann EM and Armengaud J Environ Microbiol. 2013 Aug 30 (PubMed).

Sphingomonas wittichii is an environmental bacterium that can effectively metabolize toxic pollutants such as dioxin into benign products. This study provides a nice insight into the proteins expressed by this organism in response to changes in the presence of high concentrations of pollutant molecules in the growth medium. The experiments were very consistently performed, with highly reproducible chromatographic and mass spectrometric methods applied.

This data set consisted of 8 results, generated from LC/MS/MS experiments performed on infected Macaca mulatta cells. The data files were made available through PeptideAtlas, PASS00367. It has been submitted for publication by Malouli D, Hansen SG, Nakayasu ES, Marshall EE, Hughes CM, Ventura AB, Gilbride RM, Lewis MS, Xu G, Kreklywich C, Whizin NO, Fischer MB, Legasse AW, Viswanathan K, Siess DC, Camp D, Axthelm MK, Kahl C, DeFilippis VR, Smith RD, Streblow DN, Picker LJ, and Frueh K.

The results of this study were unusual, as they are dominated by viral proteins from the infectious agent, Macacine herpesvirus 3. Herpes viruses have unusually large proteomes compared to many other types of viruses: this particular virus has 225 proteins. Many of the viral proteins were observed in most eight analyses, giving anyone interested in this virus a good starting point for comparison with their own experimental findings. The general applicability of the results was somewhat reduced by the extent of peptide carbamylation caused by the sample prepartion protocol.

For several hours today (between about 18:00 and 21:00 UTC) the main host computer for GPMDB was behaving erratically, because of an equipment problem. The problem seems to be repaired, but we will continue to monitor the system's behavior for the next few hours to be sure that the fix worked properly. Sorry for any inconvenience.

This data set consisted of 265 results, generated from LC/MS/MS experiments performed on mouse stool. The data files were made available through ProteomeXchange, PXD000240. It was published by Lichtman JS, Marcobal A, Sonnenburg JL, and Elias JE in Mol Cell Proteomics. 2013 Nov;12(11):3310-8 (PubMed).

This data set provides an excellent example of how to study a normally difficult subject: the interaction of the gut microbiome with the host-secreted gut proteome. The study uses bacterially-naive mice and populates their gut with a specific set of common human intestinal microbes to provide a model of the human digestive process. Both the proteins secreted by the mouse and those of the known prokaryotes can be easily detected, even though the original sample is less than ideal for proteomics. Any one examining the data should keep in mind that the samples contain high concentrations of host organism proteases and peptidases, resulting in numerous non-tryptic peptides that must be treated properly to obtain a good idea of the proteins present.

This data set consisted of 7 results, generated from LC/MS/MS experiments performed on yeast lysates. The data files were made available through the Chorus Project. It was published by Hebert AS, Richards AL, Bailey DJ, Ulbrich A, Coughlin EE, Westphall MS and Coon JJ in Mol Cell Proteomics. 2013 Oct 19 (PubMed).

This paper is one of a spate of recent publications demonstrating the capabilities of new experimental methods and instruments by analyzing cell lysates of lab-grown S. cerevisiae (including the L-A and L-BC viruses). In this case, the study utilizes a new mass spectrometer that is capable of taking twice as many spectra per hour as the previous generation of instrument. The results demonstrated that the additional spectra generated could be used as efficiently as the older instrument configuration, resulting in good quality peptide sequence assignments of as many as 65% of the spectra. The group make unusually good use of the chromatographic run, only losing ~10% of the run to isocratic holds and column washes. The only puzzling feature of the data was the parent ion mass accuracy: the new instrument's accuracy appears to be ± 8 ppm FWHM, compared to the older instrument's ± 2 ppm FWHM.

This data set consisted of 7 results, generated from LC/MS/MS experiments performed on samples after phospho-peptide enrichment. The data files were made available through Massive, MSV000078325. It was published by Zhuang G, Yu K, Jiang Z, Chung A, Yao J, Ha C, Toy K, Soriano R, Haley B, Blackwood E, Sampath D, Bais C, Lill JR and Ferrara N in Sci Signal. 2013 Apr 16;6(271):ra25 (PubMed).

This study provides some straightforward insights into the complexities of VEGF signalling. The experiments used iTRAQ-based quantitation and high resolution mass spectrometry to generate clean, unambiguous identifications with unusually high overall statistical significance. Any group interested in dissecting signalling involving serine-threonine or tyrosine kinases should carefully consider the methods used by the study.

This data set consisted of 29 results, generated from LC/MS/MS experiments performed on whole cell preparations. The data files were made available through ProteomeXchange, PXD000389. It was published by Shi L, Chowdhury SM, Smallwood HS, Yoon H, Mottaz-Brewer HM, Norbeck AD, McDermott JE, Clauss TR, Heffron F, Smith RD and Adkins JN in Infect Immun. 2009 77:3227-33 (PubMed).

The data associated with this study has recently been made publicly available and it provides an interesting insight into the proteins from prokaryotes that can be detected in infected eukaryote cells. In these experiments, the time course of Salmonella entrica typhimurium intracellular infection in transformed mouse macrophages was explored. Both the mouse cell proteome's response to infection and the gradual emergence of the S. enterica proteome in the cells can be tracked. Any group interested in the proteomics of infectious disease or the detection of microbiome proteomes in intact organisms would benefit from examining the results. The study also provides some very good examples of the gag:p, gag-pol:p and env:p proteins from the Moloney murine leukemia virus, one of the retroviruses necessary for the initial transformation of the mouse cell line.

This data set consisted of 30 results, generated from a combination of multidimensional chromatography and SILAC quantitation. The data files were made available through ProteomeXchange, PXD000385. It was published by Cho CK, Drabovich AP, Karagiannis GS, Martínez-Morillo E, Dason S, Dimitromanolakis A, Diamandis EP in Clin Proteomics. 2013 10:2 (PubMed).

This study was designed to determine the quantitative characteristics of the detectable proteome of amniocytes (cells floating freely in amniotic fluid) obtained from patients with Down Syndrome. The experiments were well executed and the mass spectrometry very consistently performed. The chromatographic method used resulted in unusually good recovery of hydrophobic (late eluting) peptides. The method also seemed somewhat biased against hydrophilic (early eluting) peptides, which should be kept in mind by any group reproducing these measurements.

This data set consisted of 36 results, using a novel strategy for peptide fractionation prior to LC/MS/MS. The data files were made available through ProteomeXchange, PXD000168. It was published by Han D, Moon S, Kim Y, Kim J, Jin J and Kim Y in Proteomics 2013 Aug 14 (PubMed).

These studies contain the best information to date on the proteome of the mouse BV-2 cell line, which was originally derived from microglial cells. The sample preparation methods used performed very well, producing excellent data and continuing a recent trend of protocols that lead to the recovery of phosphorylated peptides without any specific enrichment protocol. The results also showed very few experimental artifacts, reproducible chromatography and good run-to-run consistency in the high resolution MS/MS parent and fragment ion calibrations.

The 3516th daily build of GPMDB (Sept, 20, 2013) contained 1,002,098,817 peptide identifications, marking the first time GPMDB exceeded one billion entries. We would like to thank the proteomics community for making so much data obtained from so many species pubicly available, giving us the opportunity to display the results to the biomedical and scientific research communities.

This data set consisted of 452 results, consisting of LC-MS/MS runs of individual gel bands. The data files were made available through ProteomeXchange, PXD000359. It was published by Zanivan S, Maione F, Hein MY, Hernández-Fernaud JR, Ostasiewicz P, Giraudo E and Mann M in Mol Cell Proteomics 2013 Aug 26 (PubMed).

While multidimensional peptide separation technology is a very frequently used method in proteomics, much of the biomedical research community still depends on (and trusts) SDS-PAGE gel separation as their standard method of protein sample preparation. The experiments described in this manuscript show how well proteomics can perform when using gel-based separations and SILAC quantitation methods to solve a genuine biological problem. This relatively large dataset was well constructed and the associated mass spectrometry data was very consistent from run-to-run. Any group interested in the proteomics of endothelial cells should consult these results to get an idea of the proteins accessible using this type of experiment.

This data set consisted of 32 results, from bait pull-down affinity purification experiments. The data files were made available through ProteomeXchange, PXD000052. It was published by Bounab Y, Hesse AM, Iannascoli B, Grieco L, Coute Y, Niarakis A, Roncagalli R, Lie E, Lam KP, Demangel C, Thieffry D, Garin J, Malissen B and Daeron M in Mol Cell Proteomics 2013 Jul 2 (PubMed).

This set of experiments does an excellent job of demonstrating how complex affinity pull-down experiments have become because of the improved sensitivity of proteomics techniques and instrumentation. The results were very clean, with a minimum of experimental artifacts and a well-reproduced chromatographic protocol. The data also showed surprisingly good recovery of phosphopeptides, which are often lost in studies of this type unless special care is taken.

This data set consisted of 57 results, from 1D chromatographic separations coupled to LC/MS/MS. The data files were made available through ProteomeXchange, PXD000278. It was published by Liu NQ, Dekker LJ, Stingl C, Güzel C, De Marchi T, Martens JW, Foekens JA, Luider TM and Umar A. in J Proteome Res. 2013 Aug 20 (PubMed).

This study does a very good job of demonstrating which proteins could be readily identified and quantified from a straightforward laser microdissection protocol used on real tissue samples. Overall, the samples were worked up consistently and the chromatography was consistently well done. The data showed that the proteins were not chemically degraded by the laser microdissection method used and that it was a viable tissue preparation technique for proteomics studies.

We have compiled a comprehensive list of all human protein ubiquitination sites represented by good quality data in GPMDB. This list has been subdivided on a chromosome-by-chromosome basis, using ENSEMBL v. 70 as the source of the protein and gene sequences. All of the splice variants listed by ENSEMBL have been annotated.

The files associated with the annotation for each chromosome (and a merged list of all chromosomes) is now available by FTP. A description of the format of these files (README.txt) is in the same directory. A short summary of the number of ubiquitin-modified proteins, genes and sites is given here. For unique protein sequences in the proteome, the overall totals are as follows:

This data set consisted of 899 results, from experiments aimed at either broad surveys of the full proteome or kinome of all 59 cell lines and high sensitivity analysis of 9 selected cell lines. The data files were made available through Proteomics DB. It was published by Moghaddas Gholami A, Hahne H, Wu Z, Auer FJ, Meng C, Wilhelm M and Kuster B in Cell Rep. 2013 4:609-20 (PubMed).

This study did a good job of generating a large, heterogenous (but quite consistent) data set from a variety of standard Homo sapiens cell lines that have been widely used in biomedical research. Any researcher interested in a source of good quality peptide MS/MS data for almost any purpose should consider this data set carefully. The range of workflows used and the types of experiments performed were representative on current proteomics technology, with a strong emphasis on reproducibility and standardization of techniques.

This data set consisted of 12 results, from experiments using a phosphopeptide affinity enrichment workflow. The data files were made available through Chorus. It was published by Stahl JA, Chavan SS, J Sifford , Macleod V, Voth D, Edmondson RD and Forrest JC. in Plos Pathogens 2013 (in press).

The information made available in this study is the first public data set that does a good job of sketching out simultaneously the protein phosphorylation patterns in both an infectious agent (in this case murid herpesvirus 4) and the host (Mus musculus). The extent of phosphorylation of the viral proteins was significant and reproducible within the study. Hopefully this type of experiment will be performed more frequently, given the clinical importance of herpesvirus infections in mammalian hosts.

This data set consisted of 131 results, from a variety of different laboratory workflows. The data files were made available through PeptideAtlas, PASS00111. It was published by Tanca A, Biosa G, Pagnozzi D, Addis MF and Uzzau S in Proteomics 2013 Jun 20 (PubMed).

This study does a good job of providing information that can be used to understand the effects of specific experimental workflows on the outcomes of proteomics analysis. By starting with the same sample and using multiple protein extraction/fractionation methods, the resulting data sets show the details of a protein's detection can vary both quantitatively and qualitatively depending on how the sample was treated. These results should be very useful both for planning experiments and for general tutorial purposes.

This data set consisted of 247 results, generated by multidimensional chromatography and titanium dioxide affinity purification using isotopically labelled dimethylation for relative quantitation. The data files were made available through ProteomeXchange, PXD000222. It was published by Halim VA, Alvarez-Fernández M, Xu YJ, Aprelia M, van den Toorn HW, Heck AJ, Mohammed S and Medema RH. in Sci Signal. 2013 Apr 23;6(272):rs9 (PubMed).

These results provide an interesting insight into the use of a simple isotope-labelled chemical tag (N-dimethyl) as a quantitation reagent in the pursuit of a relatively rare sub-class of peptides, in this case phospho-peptides. The experiments were uniformly well done, although some artifacts (such as carbamylation) could be detected throughout. Any one interested in the analysis of data generated using this reagent for quantitation should consider these results as a very good set of examples for understanding the LOD/LOQs possible in state-of-the-art measurements.

GPMDB adds new peptide-to-spectrum assignments incrementally once daily, resulting in a new "build" (today's build is #3464). We track the number of new assignments at the peptide, protein and residue levels and make these historical numbers available by FTP. We use this information to track the health of the system and to understand seasonal fluctuations in the availability of new data.

A combination of factors has led to significant increase in the number of new peptide-to-spectrum assignments this year. Faster, higher accuracy instruments, an increased emphasis on making raw data available and the availability of easy-to-use raw data repositories have all contributed to this burst of new assignments. The histogram below shows the number of new peptide-to-spectrum assignments added in 30 day increments for the period February-July 2013.

This data set consisted of 13 results, generated by LC/MS from human embryonic stem cell cultures, with lysine/arginine SILAC for quantitation. The data files were made available through ProteomeXchange, PXD000151. It was published by Liberski AR, Al-Noubi MN, Rahman ZH, Halabi NM, Dib SS, Al-Mismar R, Billing AM, Krishnankutty R, Ahmad FS, Raynaud CM, Rafii A, Engholm-Keller K and Graumann J in J Proteome Res. 2013 Jul 5;12:3233-45 (PubMed).

The work focusses mainly on demonstrating that SILAC quantitation can be adapted to human embryonic stem cell cultures that use "feeder-free" methods (i.e., cell culture without a substrate layer of mouse embryonic fibroblasts). The data clearly supported the stated goal of the study, but it also provides a very good resource for anyone requiring good SILAC data to demonstrate the utility of an algorithm or analysis method. The results were very clean, with few contaminants and a minimum of experimental artifacts. The chromatography was well done and consistent from run-to-run.

This data set consisted of 160 results, generated by LC/MS from samples prepared using several different protein and peptide fractionation methods. The data files were made available through ProteomeXchange, PXD000231. It was published by Darville LN and Sokolowski BH in J Proteome Res. 2013 Jun 26 (PubMed).

This study exemplifies why the isolation of unusual tissues is so important in the proteomics of organisms. The investigators isolated protein fractions from murine cochlear sensory epithelium tissue and performed a series of protein and peptide fractionation studies in an effort to find low concentration proteins in this very specialized tissue. The results show an unusually large number of previously unobserved proteins, as well as many of the best observations of previously found, but rare, extracellular matrix proteins.

This data set consisted of 6 results, each one a single LC-MS/MS analysis of a membrane preparation. The data files were made available through ProteomeXchange, PXD000107. It was published by Casabona MG, Vandenbrouck Y, Attree I and Couté Y in Proteomics 2013 Jun 7 (PubMed).

While it is a commonplace statement in proteomics that "membrane" proteins are difficult to detect, this group has done an admirable job generating six very reproducible data sets of the inner membrane proteome associated with the common environmental prokaryote Pseudomonas aeruginosa. The data is top-notch, with an excellent combination of experimental workflow and chromatography that clearly demonstrates that membrane proteins can be readily observed when sufficient care is taken in their preparation.

This data set consisted of 24 results, each an LC-MS/MS analysis of a pulldown affinity purification experiment using an anti-phosphotyrosine antibody. The data files were made available through MASSIVE, MSV000077310. It was published by Hansen AM, Chaerkady R, Sharma J, Díaz-Mejía JJ, Tyagi N, Renuse S, Jacob HK, Pinto SM, Sahasrabuddhe NA, Kim MS, Delanghe B, Srinivasan N, Emili A, Kaper JB and Pandey A in PLoS Pathog. 2013 Jun;9(6):e1003403 (PubMed).

While the existence of tyrosine kinase activity in Escherichia coli has been known for some time, this data represents the first solid confirmation of protein tyrosine phosphorylation in this clinically important prokaryote. The data also provides evidence for the range of proteins exhibiting tyrosine phosphorylation in two strains, both the commonly used lab strain E. coli K12 str. MG1655 as well as the pathogenic strain E. coli O157:H7 str. EDL933. The level of phosphotyrosine enrichment is somewhat less than would be normally found in similar eukaryote-based experiments, but the signals very clearly demonstrate the existence of this modification in this species.

This data set consisted of 75 results comprised mainly of affinity pull-down experiments using conventional LC-MS/MS. The data files were made available through ProteomeXchange, PXD000208. It was published by Joshi P, Greco TM, Guise AJ, Luo Y, Yu F, Nesvizhskii AI and Cristea IM in Mol Syst Biol. 2013 Jun 11;9:672 (PubMed).

This study provides clear insight into the benefits of affinity purification in proteomics. Many of the proteins represented in this data set are the best observations of those protein species to date. The measurements are well-done, with very consistent chromatography and mass spectrometry techniques employed. These spectra should be of interest to anyone interested in the post-translational modifications associated with HDAC proteins.

As part of our contribution to the Human Proteome Project, we have compiled a comprehensive list of all human protein lysine acetylation sites represented by good quality data in GPMDB. This list has been subdivided on a chromosome-by-chromosome basis, using ENSEMBL v. 70 as the source of the protein and gene sequences. All of the splice variants listed by ENSEMBL have been annotated.

The files associated with the annotation for each chromosome (and a merged list of all chromosomes) is now available by FTP. A description of the format of these files (README.txt) is in the same directory. A short summary of the number of lysine-acetylated proteins, genes and sites is given here. For unique protein sequences in the proteome, the overall totals are as follows:

As part of our contribution to the Human Proteome Project, we have compiled a comprehensive list of all human protein N-terminal acetylation sites represented by good quality data in GPMDB. This list has been subdivided on a chromosome-by-chromosome basis, using ENSEMBL v. 70 as the source of the protein and gene sequences. All of the splice variants listed by ENSEMBL have been annotated. Protein N-terminal acetylation is the result of a multienzyme system that is responsible for processing nascent polypeptide chains as they emerge from the ribosome (see Thomas Arnesen's interesting review for details).

The files associated with the annotation for each chromosome (and a merged list of all chromosomes) is now available by FTP. A description of the format of these files (README.txt) is in the same directory. A short summary of the number of NT-acetylated proteins, genes and sites is given here. For unique protein sequences in the proteome, the overall totals are as follows:

This data set consisted of 124 results comprised mainly of metal-oxide affinity purified samples run using conventional LC-MS/MS. The data files were made available through PeptideAtlas, PASS00138. It was published by Courcelles M, Frémin C, Voisin L, Lemieux S, Meloche S, and Thibault P in Mol Syst Biol. 2013 May 28;9:669 (PubMed).

This well-done study is the source of a significant amount of new data associated about Rattus norvegicus phosphopeptides. Even though the rat is a common model species, there have been few phosphorylation studies using rat cell lines in which the raw data has been made publicly available compared to the large volume of human and mouse studies that have been made public. The data provided here was very good quality and should provide a useful starting point for any group interested in comparing the patterns of phosphopeptide detection across mammalian model species.

This data set consisted of 51 results comprised of multidimensional chromatography run summaries. The data files were made available through PeptideAtlas, PASS00230. It was published by Wu L, Candille SI, Choi Y, Xie D, Jiang L, Li-Pook-Than J, Tang H and Snyder M. in Nature. 2013 May 15 (PubMed).

One of the milestone works on lymphocyte protomics has been the 2007 work Lingfen Wu, et al., "Global survey of human T leukemic cells by integrating proteomics and transcriptomics profiling" (PubMed). This new work (also with Lingfen Wu as a lead author) combines state-of-the-art RNASeq measurements with very consistent deep proteomics data from cell lines derived by human herpesvirus 4 transformation of lymphoblasts from the individuals used in the HapMap project. This study has enough data to make a good run at understanding the relationship between RNA and protein detectability using some of the best instrumentation for detecting both types of macromolecule.

This data set consisted of 135 results comprised of multidimensional chromatography runs and experiment summaries. The data files were made available through the the authors' web site. It was published by Webb KJ, Xu T, Park SK and Yates JR 3rd in J Proteome Res. 2013 12:2177-84 (PubMed).

This study provides a good insight into the proteins, peptides and post-translational modifications that can be readily accessed in S. cerevisiae and its associated viruses (L-A and L-AB) under several different physiological conditions using two different multidimensional chromatography protocols. The data was of good quality and nicely demonstrated the effects of the temporal details of chromatographic conditions on spectrum identifications rates. The results also indicated the degree to which phosphorylations can now be detected in routine analysis.

This data set consisted of 28 results that comprised a single multidimensional chromatography experiment. The data files were made available through PASSEL (PASS00215). It was published by Sheynkman GM, Shortreed MR, Frey BL and Smith LM in Mol Cell Proteomics 2013 Apr 29 (PubMed).

This data set defines the state-of-the-art with respect to "deep" proteomics of a human cell line (Jurkat cells). The combination of a first dimension using high pH HPLC followed by low pH HPLC produced a very well separated collection of peptides. The use of HCD coupled with high resolution fragment ion measurements using an Orbitrap lead to very high confidence peptide assignments. Anyone interested in detecting relatively rare post-translational modifications or determining splice variants would be well served by performing their analysis on this data set first.

In addition to the human phosphorylation annotation released on Sunday, we have also prepared annotation in the same format for a set of model species commonly used in proteomics experiments. Annotation for the following species is now available: C. elegans, D. melanogaster, M. musculus and S. cerevisiae.

As part of our contribution to the Human Proteome Project, we have compiled a comprehensive list of all human protein phosphorylation sites represented by good quality data in GPMDB. This list has been subdivided on a chromosome-by-chromosome basis, using ENSEMBL v. 70 as the source of the protein and gene sequences. All of the splice variants listed by ENSEMBL have been annotated.

The files associated with the annotation for each chromosome (and a merged list of all chromosomes) is now available by FTP. A description of the format of these files (README.txt) is in the same directory. A short summary of the number of phospho-proteins, genes and sites is given here. For unique protein sequences in the proteome, the overall totals are as follows:

This data set consisted of 136 results composed of individual SDS-PAGE gel slices and experiment summaries. The data files were made available through ProteomeXchange (PXD000120). It was published by Fanayan S, Smith JT, Lee LY, Yan F, Snyder M, Hancock WS and Nice E in J Proteome Res. 2013 Mar 13 (PubMed).

The data reported here was a good example of what can be done with whole cell lysates analyzed using SDS-PAGE protein separations and low resolution (LTQ) mass spectrometry. The experiments elucidate an interesting biological issue: "How different were the protein concentrations in three related cell lines and how were those changes generated by differences in RNA concentration?" This data would be useful for anyone interested in practical difficulties associated with combining protein molecular mass information with peptide identifications when using SDS-PAGE gels for protein separations.

As some readers may have noticed, the logo for GPM and GPMDB has changed recently (thanks to noted electronic artist KD Thornton). This change is part of a general redesign of the site to conform to more modern web page coding trends, simplify page navigation and improve the usefulness of the overall site on smaller screens and mobile platforms. If you have any suggestions regarding things you would like to see in a new design (or things that really bug you about the current one), please let us know at contact@thegpm.org.

The definition of the GPMDB REST interface has been expanded to include a new method, allowing the rapid calculation of peptide ω frequencies for any set of peptides and protein accession numbers stored in GPMDB. These frequencies are useful when comparing observed peptides to those previously observed: a technical definition is given here. The description of this method and an example have been added to the GPMDB Wiki page for the REST interface.

Links to the "Protein Abundance Across Organisms" project (PaxDB) have been added to the protein-specific display pages in GPM and GPMDB for appropriate species. These links can be found by clicking the green "Protein" links button at the top of the appropriate pages. The PaxDB system (Wang, M. et al. Mol Cell Proteomics 2012, doi:10.1074/mcp.O111.014704) is, in the words of the developers:

This data set consisted of 30 results, each representing a sub-cellular fractionation experiment analyzed by reversed-phase HPLC. The data files were made available through PASSEL (PASS00190). It was published by Ori A, Banterle N, Iskar M, Andrés-Pons A, Escher C, Khanh Bui H, Sparks L, Solis-Mezarino V, Rinner O, Bork P, Lemke EA, and Beck M in Mol Syst Biol. 2013 Mar 19; 9:648 (PubMed).

This study represents a thorough examination of the proteins associated with an important intracellular structure, the human nuclear pore. A variety of methods were applied to the problem, such as protein chemistry isolations, proteomics, high resolution optical and electron microscopies, and the study tries to synthesize all of these measurements together to obtain a model of how nuclear pores differ between common cell lines (SK-MEL5, RKO, HEK 293 and HeLa). The proteomics data was excellent and demonstrates how well these techniques can be used as a component in the examination of a complex problem.

This data set consisted of 114 results, each representing an affinity-purification experiment analyzed by reversed-phase HPLC. The data files were made available through PASSEL (PASS00226). It was published by Varjosalo M, Keskitalo S, Van Drogen A, Nurkkala H, Vichalkovski A, Aebersold R and Gstaiger M in Cell Reports 2013 25 April, 3(4):1306–20 (PubMed).

The data presented with this manuscript will give any interested investigator significant insight into the work necessary to turn a set of identifications into a high-quality protein-protein interaction map. The proteomics experiments were consistently done, with attention to the often overlooked chromatographic details. Anyone interested in performing or analyzing this type of information would be well served by examining these experiments prior to planning their own. One feature of the data not commented on in the manuscript were the clear signals for Human adenovirus C E1A and E1B proteins (these proteins are coded on the viral DNA responsible for the original HEK 293 transformation). These proteins are commonly observed in HEK 293 proteomics data, but the presence of these proteins in specific pull-down experiments here may shed some light on their role in CMGC kinase signalling, which to our knowledge has not yet been examined.

This data set consisted of 33 results, each representing chemical modifications, pull-downs and reversed-phase HPLC peptide separations. The data files were made available through ProteomeXchange (PXD000141). It was published by Wolfe LM, Veeraraghavan U, Idicula-Thomas S, Schurer S, Wennerberg K, Reynolds R, Besra GS, and Dobos KM in Mol Cell Proteomics. 2013 Mar 5 (PubMed).

This data represents a particularly interesting class of proteomics experiments: the combination of specific chemical probes that facilitate protein and peptide purification techniques that very specifically target particular pathways and/or biochemistry. In this case, it is the incorperation of desthiobiotin into proteins that bind ATP. This type of experiment does not require cutting-edge high throughput instruments, rather its value is associated with developing a coherent model of how well the modification reagent is functioning in the experimental system and then evaluating the results in the context of that model and the known biochemistry of the system.

This data set consisted of 191 results, each representing a combination of peptide separation methods and reversed-phase HPLC peptide separations. The data files were made available through PASSEL (PASS00211). It was published by Hassan C, Kester MG, Ru AH, Hombrink P, Drijfhout JW, Nijveen H, Leunissen JA, Heemskerk MH, Falkenburg JH, and Veelen PA in Mol Cell Proteomics. 2013 Mar 19 (PubMed).

This study was an interesting investigation of how to purify and detect a special class of endogenous peptides: those presented on the surface of leukocytes by human leukocyte antigen class I molecules. These peptides are relatively short (~9 residues) and are generated by ubiquitin-mediated proteolysis in the cell's proteasome. The peptides are also present at relatively low concentrations in normal cells and since they have no special properties, they can be difficult to isolate. The protocols described in this paper do a good job of purifying these peptides and demonstrated the merits of different peptide separations methods when applied to this situation. Any group interested in the large-scale analysis of HLA class I peptides should examine the data in this work carefully.

This data set consisted of 384 results, each representing a combination of protein separation methods and reversed-phase HPLC peptide separations. The data files were made available through PASSEL (PASS00069). It was published by Gunaratne J, Schmidt A, Quandt A, Neo SP, Sarac OS, Gracia T, Loguercio S, Ahrne E, Li Hai Xia R, Tan KH, Loessner C, Bahler J, Beyer A, Blackstock W, and Aebersold R in Mol Cell Proteomics. 2013 Mar 5 (PubMed).

The experiments described in this paper were an attempt to characterize the proteome of the yeast Schizosaccharomyces pombe using modern instruments and standard sample preparation methods. The laboratory model organism S. pombe is descended from an environmental species and it has been extensively studied for determining the mechanisms behind its cell cycle and genetics. S. pombe has been less studied by proteomics methods than its other beer-related relative, S. cerevisiae, and this study does a lot to address that deficit.

Yesterday (April 2, 2013) marked the tenth anniversary of the first public release of the proteomics search engine X! Tandem. At the time, it was the first open source protein identification software project to become available (a previous academic effort had been attempted but was shutdown because of patent concerns). The release of the X! Tandem code – and the subsequent release a few months later of the NCBI-sponsored open source search engine OMSSA – ushered in the modern era of proteomics informatics and computational biology.

This data set consisted of 382 results, each representing a fraction from multidimensional chromatography experiments, using iTRAQ derivatives for quantitation. The data files were made available through PASSEL (PASS00056). It was published by Rose CM, Venkateshwaran M, Volkening JD, Grimsrud PA, Maeda J, Bailey DJ, Park K, Howes-Podoll M, den Os D, Yeun LH, Westphall MS, Sussman MR, Ané JM, and Coon JJ in Mol Cell Proteomics 2012 11:724-44 (PubMed).

This study represented the first major effort to characterize the proteome of the model plant species Medicago truncatula, a legume species in the same genus as the commercially important forage crop Medicago sativa. The study takes advantage of the known genome of the organism and provides some interesting insights into the role of protein phosphorylation in metabolic processes associated with interaction between the plant and the symbiotic nitrogen-fixing bacteria that it hosts in root nodules. The data was of excellent quality and it provides the best resource available for obtaining phosphodomain annotation for legume proteins.

This data set consisted of 288 results, each representing a single affinity purification experiment. The data files were made available through PASSEL (PASS00117). It was published by Varjosalo M, Sacco R, Stukalov A, van Drogen A, Planyavsky M, Hauri S, Aebersold R, Bennett KL, Colinge J, Gstaiger M and Superti-Furga G in Nat Methods 2013 Mar 3 (PubMed).

These experiments represent probably the best resource available for data that can be used to teach the practice of computational biology as it applies to proteomics. The raw data, as well as the information generated by two different peptide identification systems, can be directly downloaded and used. The preliminary analyses available (on which the manuscript was based) provide a nice insight into the current generic practices used for peptide and protein identification by many laboratories. Given the data and information available in these files, it is practical to create a series of tutorial subsets of the files that can be manipulated directly by students to answer many questions relevant to practical proteomics data analysis, for example:

This data set consisted of 220 results consisting of multidimensional chromatography single HPLC runs and experiment summaries. The data files were made available through Proteomexchange (PXD000016). It was published by Haussmann U, Wolters DA, Fraenzel B, Eltis LD, and Poetsch A. in J Proteome Res. 2013 12:1188-98 (PubMed).

The organism being studied here (Rhodococcus jostii) belongs to a genus of largely non-pathogenic prokaryotes belonging to the large suborder Corynebacterineae that contains Mycobacteria and Corynebacteria. It is known to be very good at utilizing a wide variety of substrates for growth, including many types of aromatic compounds that are not easily metabolized by other bacteria. This study does an excellent job of characterizing changes in the membrane proteins present in the organism in response to changing its carbon source from mainly pyruvate to cholesterol and cholate. The experiments were consistently well done and the results provide an excellent resource for investigating how well membrane proteins can be recovered using standard proteomics techniques. An interesting feature of the results is the apparent presence of a threonine/serine specific, protein-N-acetyl transferase, which results in protein N-terminal acetylation at threonine residues (with reduced activity for serine).

The latest version of X! Tandem 2012.02.01 has completed its on-line testing and should be considered a stable release. This new version of the open source search engine has several new features, including enhanced compatibility with the SRM/MRM project Skyline (thanks to Brendan Maclean and Brett Phinney). It also has a new sequence stacking mechanism that effectively produces non-redundant sequence databases on the fly, which can significantly improve performance when using the proteomes of multiple strains of prokaryote species or eukaryote protein sequences that correspond to multiple alternate splice variants that are only different in the mRNA's UTRs.

The main GPM system has been updated to use the latest version of the human proteome — ENSEMBL v. 70.37 — which was based on the human genome sequence GRCh37.p10, Feb 2009. All of the relevant resources (including annotated spectral library and proteotypic peptides) have been updated to the new sequence set. The annotation file for human SNAPs (Single Nucleotide Amino acid Polymorphisms) has been updated to dbSNP 137 (1,690,969 SNAPs), using ENSEMBL's Biomart interface for the DNA-to-protein coordinate and allele mapping.

The Journal of Proteome Research has released it's special issue associated with the launch of the chromosome-based Human Proteome Project. This issue is freely available on-line and it includes articles from many of the cHPP national groups, as well as technical papers associated with the creation of boutique databases, laboratory techniques and visualization systems specifically designed for use by the cHPP. There are also several general articles on the current status of experimental evidence for the existence of human proteins, discussing the gaps in our knowledge at the moment.