Systematic Datasets Research Articles

実油田シミュレーションモデル上でのニューラルネットワークを用いた原油生産計画の最適設計

This paper investigates the applicability of Artificial Neural Networks (ANN) in the optimization process of lifting rate allocation for ESP wells. In the optimization, oil recovery is maximized according to the production performance predicted using a reservoir simulation of a real oil field.Lufeng 13-1 Oil Field in the South China Sea has strong natural bottom-water drive and its development has reached a mature production stage. All production wells are currently operated with Electric Submersible Pumps (ESP) and the averaged water cut has exceeded 90%. The limited handling capacity of the surface processing facilities cannot allow all the wells to lift liquid at their ESPs' maximum rate. The lifting rate allocation to the wells, therefore, needs to be optimized for maximizing the oil recovery in a certain production period.Generally, a large number of reservoir simulation runs are required for solving this type of optimization problems since all possible cases of lifting rate allocations need to be examined. In this study, oil recovery volume was estimated with a trained ANN on the basis of allocated liquid rates at the wells as the initial screening. The ANN was constructed with back-propagation method treating the results of 100 simulation results as the training data sets. Two patterns of training data sets were examined; one with random rate allocation and the other with systematic rate allocation. Blind tests for the estimation accuracy presented that the ANN trained with the systematic data sets showed better results than that with the random ones. Oil recovery factors under all the possible cases with different liquid rate combinations were estimated using the trained ANN. The top fifty cases were selected for the final examination by numerical simulation. The best case yielded 5.5% increase of produced oil volume from the base case, in which lifting rates were equally reduced to 89% of the maximum lifting capacities to meet the facility capacity. The investigation results demonstrated that the efficiency of the optimization was remarkably improved with the use of ANN on the determination of optimum liquid lifting rates in terms of oil recovery.

Assessment of naturally occurring radionuclides around England and Wales: Application of the G-BASE dataset to estimate doses to non-human species

An assessment of the natural radiation dose to wildlife in England and Wales was made to determine the contribution it may make to the total radiation dose estimated during environmental impact assessments. Significant use was made of systematic datasets for environmental media (stream sediments, stream waters and soils), in particular those produced by the Geochemical Baseline Survey of the Environment (G-BASE) project. This provided extensive, although variable, coverage for different elements and sample types after normalisation of data to account for changes in sampling and analysis over time. Almost complete coverage for K in stream sediments was achieved by merging G-BASE andWolfson Atlas data. This required normalisation of theWolfson data to the G-BASE results. Coverage was improved greatly for U and Th in sediments, and K, U and Th in soils, by using the strong relationship between soils and sediments and geology (both solid and superficial) to extrapolate the data. The total U, Th and K data were used to derive activity concentrations of 238U and 232Th series radionuclides and 40K. External dose rates to wildlife were then estimated from derived media concentrations; internal dose rates were estimated from measured activities in biota or activities predicted using recommended concentration ratios.

Unconscious Applicants: A Systematic Test of the Name-Letter Effect

It is now generally accepted that unconscious processesmodulate human behavior. Nelson and Simmons (2007) provided a thoughtprovoking illustration of this phenomenon. They demonstrated that people’s unconscious preferences for the letters in their own names are strong enough that people pursue name-related performance outcomes that they would consciously avoid. For instance, although most students consciously strive to earn As, students whose names begin with C andD have lower grade point averages (GPAs) than do students whose names begin with A and B. These findings build on previous research on the name-letter effect (NLE), which is the tendency to evaluate letters in one’s name more favorably than other letters. The influence of this implicit preference is striking: It affects important decisions, such as where people choose to live and whom they decide to marry (Pelham, Carvallo, & Jones, 2005). However, NLE studies have not gone unchallenged. As did previous work, Nelson and Simmons’s (2007) study suffered from a number of limitations that may compromise its interpretation as demonstrating the NLE. We conducted a large-scale study to investigate whether people whose names begin with a given letter have a tendency to work for companies with names that begin with the same letter. Our study addresses the limitations of previous work and extends the existing NLE literature in four ways. First, researchers have typically sampled a few letters or names to test for nameinitial-congruent outcomes (e.g., is Jack more likely to live in Jacksonville than in Philadelphia; see Pelham, Mirenberg, & Jones, 2002). As Gallucci (2003) argued, these tests ‘‘can yield significant results due to spurious effects and sampling biases’’ (p. 789), so interpretation of the results as demonstrating the NLE is compromised. Even (or especially) when specific cases are sampled from very large databases (e.g., when the frequencies of two names in two cities are compared in a 2 2 table; Pelham et al., 2002), findings may be an artifact of ‘‘luck’’ in choosing two of the many possible names or letters for testing. The first four of five studies by Nelson and Simmons (2007) also had this problem: For instance, when testing for an NLE on GPAs, they compared only students with the initials A and B against those with the initials C and D. However, although students with the initials C and D received somewhat lower GPAs (i.e., on average, approximately 3.35, vs. 3.37 for students with initials A and B), such small variations might also have been observed for other initials. Maybe students with the initial M also had lower GPAs than students whose names began with A or B. In Nelson and Simmons’s elegant fifth study, NLEs were obtained with a wider range of initials. Unfortunately, these findings were constrained by the limitations of the laboratory (e.g., results may have been affected by demand characteristics) and lack ecological validity. The study we report in this article offers a convincing supplement to the NLE literature because the design encompassed all letters in a very large sample of realworld behavior: job choice. Thus, our study was not susceptible to the risk of sampling bias present in previous operationalizations of the NLE. Second, it is important theoretically to establish the boundary conditions for psychological phenomena, particularly when they are provocative and surprising. The best way to explore these boundaries for the NLE is to put them in direct competition with behavior driven by conscious thought. Rational and deliberate thinking processes guide job choice: Applicants are attracted by objective job and organizational attributes, such as salary, job content, and location (Chapman, Uggerslev, Carroll, Piasentin, & Jones, 2005). Therefore, job choice is a theoretically challenging case for the NLE. Third, although researchers have claimed that NLEs are stronger for less frequent letters and names, which are more distinctive (Pelham et al., 2005), tests of this hypothesis have not confirmed it. Jones, Pelham, Mirenberg, and Hets (2002) even found a numerical tendency toward larger NLEs for names with more frequent initials. The design of the present study allowed a more comprehensive test of this hypothesis, so that the study might contribute to a better theoretical understanding of the mechanisms involved in the NLE. Fourth, given the lack of large and systematic data sets with accurate a priori frequency estimates, reliable estimates of the size of the NLE on real-world behavior are not available. From a conceptual standpoint, such estimates are crucial for comparing Address correspondence to Frederik Anseel, Department of Personnel Management, Work and Organizational Psychology, Ghent University, Henri Dunantlaan 2, B-9000 Ghent, Belgium, e-mail: frederik.anseel@ugent.be. PSYCHOLOGICAL SCIENCE

DNA barcoding demystified

Only 10% of the earth’s biota has been described despite250 yearsoftaxonomicresearch(Wilson2000).Thisisinlargepart a reflection of the extent and complexity of biologicaldiversity,butitisalsotruethattraditionaltaxonomictechniquesare labourious and highly specialised, and taxonomic expertiseis very thinly spread across the myriad groups of life (Scotland

Particle holdup in a liquid–solid circulating fluidized bed

AbstractAn experiment was conducted to acquire a set of systematic data of particle holdup in risers of a liquid–solid circulating fluidized bed. In the experiment, two kinds of riser were provided, their inner diameter being 24 mm and 36 mm, respectively. Tested particles were of glass and ceramics, having a diameter range from 2.10 to 4.95 mm. Water at ambient conditions was used as the fluidizing liquid. Particle holdup was measured using a shut‐off method. Based on the experimental data, a correlation for predicting the particle holdup was derived, which could reproduce almost all experimental data with an accuracy of ±15%. The effect of the wall was not recognized within the experimental range, i.e., the diameter ratio of particle to riser is less than 0.2. The independent parameters affecting the flow characteristics of liquid–solid circulating fluidized beds were identified. © 2008 Wiley Periodicals, Inc. Heat Trans Asian Res, 37(3): 184–196, 2008; Published online in Wiley InterScience (www. interscience.wiley.com). DOI 10.1002/htj.20194

The power of indicators: introduction to special issue on public project funding of research

HIS SPECIAL ISSUE presents the results of a number of studies that have been undertaken in the PRIME Network of Excellence and in ERAWATCH to characterise and compare public funding of research activities and, more specifically, project funding, broadly meaning funds allocated to research teams from external agencies to perform research activities limited in time, scope and budget. The relevance of this issue in the framework of today’s public debate on how to promote research activities at the national and European level is very clear, as well as the need to dispose of systematic comparisons of the different national funding systems and of their evolution over time. Also, there is a large body of literature in science policy studies dealing with the organisation of public research funding and debating the advantages and disadvantages of the models and organisational structures; sufficient to mention the debate concerning the need to provide enough support to basic investigatordriven research and to avoid too unilateral an orientation towards immediately useful research. While building on this stock of knowledge and the conceptual models and categories developed by a number of authors (for which we refer to the theoretical part of the papers), our work has been based on an important methodological innovation — the attempt systematically to produce a set of descriptors and indicators to compare funding systems quantitatively across countries and time. Since R&D statistics largely disregard the issue of characterising the allocation channels of public money, we have devised a specific methodology based on the collection of data from public budgets and funding agencies, and then their elaboration to provide comparative indicators. At the same time, project participants had to provide systematic and standard descriptions of the funding instruments, because we knew from previous work that the lack of this information in a usable form greatly impaired comparative work. This methodological work took almost two years and is presented in detail in a companion paper published in Research Evaluation (Lepori et al, 2008a): the reader should thus refer to it to understand the background of the figures presented here. It is, however, important to understand correctly the difference (and complementarity) between this work and the development of R&D statistics as undertaken by official bodies. While the work of official bodies aims to construct a systematic set of data concerning research expenditures in a long-term and comparative perspective by collecting original data through surveys (Godin, 2005), our aim was more modest. It was to provide, through the ad hoc elaboration of existing data, the answer to some very specific and localised questions, such as comparing across countries the portfolio of public funding or measuring the evolution over time of its share in total public funding. Hence the indicators proposed in these papers are good examples of what we called ‘positioning indicators’, that is, ones aiming to characterise the position and linkages of the different actors in the national innovation systems (Lepori et al, 2008b). A second relevant feature of these indicators is that they are not meant to provide complete descriptions of the reality on their own, nor to be used T

Ozone in Spain's National Parks and Protected Forests

In general, it is difficult to measure air pollutant concentrations in remote areas, as they are mostly national parks and protected areas. Passive samplers provide an accurate and inexpensive method for measuring cumulative exposures of different air pollutants. They have been used to collect ozone data in both laboratory and field at different geographical scales. The objective of the present study is to fill the knowledge gap regarding air quality in remote areas of Spain, such as national parks and protected areas. Because there were no systematic data sets on the main air pollutants that could affect these areas, an air quality measurement network was established between 2001 and 2004 on 19 locations inside Spanish national parks and protected areas. The data collected suggest that ozone levels in mountainous areas are high enough to affect sensitive vegetation. Most of the locations registered moderate-to-high ozone levels, with important interannual variability. Altitudinal ozone gradients were observed in most of the parks with complex topography due to the establishment of local circulations that incorporate polluted air masses from polluted airsheds or even long-range transport (i.e., Canary Islands). Different latitude-dependent, yearly cycles were also observed, showing two, one, or no clear peaks depending on the region. These findings extend to the most southerly locations, except in the Canary Islands, where pollution transported from other regions in the upper transport layers probably led to the high concentrations observed.

Correlation between nanostructural parameters and conductivity properties for molecularly-mediated thin film assemblies of gold nanoparticles

The understanding of the particle size and interparticle spatial effects on the electrical conductivity properties of molecularly mediated thin film assemblies of nanoparticles is important for exploring the unique electrical properties in applications such as chemical sensors and biosensors. This paper reports findings of an investigation of such effects for thin film assemblies of gold nanoparticles of highly-monodispersed sizes (2–4 nm) using alkyl dithiols of different chain lengths (0.8–2 nm). The conductivities of the thin films were measured using interdigitated microelectrodes as the platform. Experimental results have shown that the activation energy increases with chain length and decreases with particle size, and the electron tunneling decay term decreases with particle size. The results have revealed that the experimentally determined conductivity and activation energy data for the nanoparticle films quantitatively match the calculations from electrostatic model of granular metals through the electron tunneling mechanism. The strong correlation between the experimental and calculated data was attributed to a combination of the high monodispersity of the nanoparticles and the uniformity and stability of the thin film assemblies. These findings have provided an important set of systematic data supporting the applicability of the activated electron tunneling theory to the molecularly-mediated thin film assemblies of gold nanoparticles, which have important implications for the design and fine-tuning of nanostructured thin films as chemical and biological sensing materials.

Qualitative network models and genome-wide expression data define carbon/nitrogen-responsive molecular machines in Arabidopsis

Qualitative network models and genome-wide expression data define carbon/nitrogen-responsive molecular machines in Arabidopsis and indicate that regulation by carbon/nitrogen metabolites occurs at multiple levels.

European phenological response to climate change matches the warming pattern

AbstractGlobal climate change impacts can already be tracked in many physical and biological systems; in particular, terrestrial ecosystems provide a consistent picture of observed changes. One of the preferred indicators is phenology, the science of natural recurring events, as their recorded dates provide a high‐temporal resolution of ongoing changes. Thus, numerous analyses have demonstrated an earlier onset of spring events for mid and higher latitudes and a lengthening of the growing season. However, published single‐site or single‐species studies are particularly open to suspicion of being biased towards predominantly reporting climate change‐induced impacts. No comprehensive study or meta‐analysis has so far examined the possible lack of evidence for changes or shifts at sites where no temperature change is observed. We used an enormous systematic phenological network data set of more than 125 000 observational series of 542 plant and 19 animal species in 21 European countries (1971–2000). Our results showed that 78% of all leafing, flowering and fruiting records advanced (30% significantly) and only 3% were significantly delayed, whereas the signal of leaf colouring/fall is ambiguous. We conclude that previously published results of phenological changes were not biased by reporting or publication predisposition: the average advance of spring/summer was 2.5 days decade−1 in Europe. Our analysis of 254 mean national time series undoubtedly demonstrates that species' phenology is responsive to temperature of the preceding months (mean advance of spring/summer by 2.5 days°C−1, delay of leaf colouring and fall by 1.0 day°C−1). The pattern of observed change in spring efficiently matches measured national warming across 19 European countries (correlation coefficient r=−0.69, P<0.001).

Genomics in C. elegans: So many genes, such a little worm: Table 1.

The Caenorhabditis elegans genome sequence is now complete, fully contiguous telomere to telomere and totaling 100,291,840 bp. The sequence has catalyzed the collection of systematic data sets and analyses, including a curated set of 19,735 protein-coding genes--with >90% directly supported by experimental evidence--and >1300 noncoding RNA genes. High-throughput efforts are under way to complete the gene sets, along with studies to characterize gene expression, function, and regulation on a genome-wide scale. The success of the worm project has had a profound effect on genome sequencing and on genomics more broadly. We now have a solid platform on which to build toward the lofty goal of a true molecular understanding of worm biology with all its implications including those for human health.

Axial nonuniformity of geometric and mechanical properties of mouse aorta is increased during postnatal growth

The hemodynamic conditions of aorta are relatively uniform prenatally and become more heterogeneous postnatally. Our objective was to quantify the heterogeneity of geometry and mechanical properties during growth and development. To accomplish this objective, we obtained a systematic set of data on the geometry and mechanical properties along the length of mouse aorta during postnatal development. C57BL/6 mice of ages 1-33 days were studied. The ascending aorta was cannulated in situ and preconditioned with several cyclic changes in pressure. We investigated the axial variations of geometry (diameter and length) and mechanical properties (stress-stain relation, elastic modulus and compliance) of the mouse aorta from the aortic valve to the common iliac. Our results show that the arterial blood pressure of mice increased from approximately 30 to 80 mmHg during the first 2 wk of life. The stretch ratio, diameter, wall (intima-media) thickness, and total lumen volume of mouse aorta increased with age. The aorta was transformed from a cylindrical tube at birth to a tapered structure during growth. Furthermore, we found the mechanical properties were fairly uniform along the length of the aorta at birth and become more nonuniform with age. We conclude that the rapid change of blood pressure and blood flow after birth alter the geometric and mechanical properties differentially along the length of the aorta. Hence, the axial nonuniformity of the aorta increases as the organ becomes more specialized during growth and development.

A Compendium of Signals and Responses Triggered by Prodeath and Prosurvival Cytokines

Cell-signaling networks consist of proteins with a variety of functions (receptors, adaptor proteins, GTPases, kinases, proteases, and transcription factors) working together to control cell fate. Although much is known about the identities and biochemical activities of these signaling proteins, the ways in which they are combined into networks to process and transduce signals are poorly understood. Network-level understanding of signaling requires data on a wide variety of biochemical processes such as posttranslational modification, assembly of macromolecular complexes, enzymatic activity, and localization. No single method can gather such heterogeneous data in high throughput, and most studies of signal transduction therefore rely on series of small, discrete experiments. Inspired by the power of systematic datasets in genomics, we set out to build a systematic signaling dataset that would enable the construction of predictive models of cell-signaling networks. Here we describe the compilation and fusion of approximately 10,000 signal and response measurements acquired from HT-29 cells treated with tumor necrosis factor-alpha, a proapoptotic cytokine, in combination with epidermal growth factor or insulin, two prosurvival growth factors. Nineteen protein signals were measured over a 24-h period using kinase activity assays, quantitative immunoblotting, and antibody microarrays. Four different measurements of apoptotic response were also collected by flow cytometry for each time course. Partial least squares regression models that relate signaling data to apoptotic response data reveal which aspects of compendium construction and analysis were important for the reproducibility, internal consistency, and accuracy of the fused set of signaling measurements. We conclude that it is possible to build self-consistent compendia of cell-signaling data that can be mined computationally to yield important insights into the control of mammalian cell responses.

Investigation of Federal Bureau of Investigation Bomb-Related Death and Injury Data in the United States between 1988 and 1997

The study of physical injury from terrorist explosives is an increasing international area of research. However, there are few data sets to characterize the scope of injury and death from these devices. Therefore, one option is to begin evaluating statistics reported by a nontraditional public health data source, the U.S. Federal Bureau of Investigation (FBI) Bomb Data Center. We reviewed data reported by the FBI Bomb Data Center for the years 1988-1997 and analyzed the number of bomb-related deaths and injuries and incidence of bombings. The FBI reported 17,579 bombings, 427 related deaths, and 4,063 injuries in the United States between 1988 and 1997. The benefits of this data are reporting of information not normally found in public health data, including type of explosive device and explosive composition. The primary limitations include lack of case comparison and unknown methods of data reporting and data collection. To completely study physical injury from explosive devices requires a systematic and comprehensive data set. The FBI data provides an interesting statistical resource to assess the scope of injury from bombs in the United States, but at the current time cannot be used for extensive epidemiological analysis.

Pocketome via Comprehensive Identification and Classification of Ligand Binding Envelopes

We developed a new computational algorithm for the accurate identification of ligand binding envelopes rather than surface binding sites. We performed a large scale classification of the identified envelopes according to their shape and physicochemical properties. The predicting algorithm, called PocketFinder, uses a transformation of the Lennard-Jones potential calculated from a three-dimensional protein structure and does not require any knowledge about a potential ligand molecule. We validated this algorithm using two systematically collected data sets of ligand binding pockets from complexed (bound) and uncomplexed (apo) structures from the Protein Data Bank, 5616 and 11,510, respectively. As many as 96.8% of experimental binding sites were predicted at better than 50% overlap level. Furthermore 95.0% of the asserted sites from the apo receptors were predicted at the same level. We demonstrate that conformational differences between the apo and bound pockets do not dramatically affect the prediction results. The algorithm can be used to predict ligand binding pockets of uncharacterized protein structures, suggest new allosteric pockets, evaluate feasibility of protein-protein interaction inhibition, and prioritize molecular targets. Finally the data base of the known and predicted binding pockets for the human proteome structures, the human pocketome, was collected and classified. The pocketome can be used for rapid evaluation of possible binding partners of a given chemical compound.

Seasonal Variability of Cyclone Heat Potential in the Bay of Bengal

Monthly maps of cyclone heat potential (CHP) in the Bay of Bengalhave been prepared by using Levitus climatological data set. Seasonal variability ofCHP in the Bay of Bengal has been studied using the CTD data sets collected duringfive cruises during the period, 1993–1996. High value (>30 kcal/cm2) of CHP coincided with anticyclonic gyre (ACG) and the low value of CHP (16 kcal/cm2) coincided with thecyclonic gyre (CG). This emphasizes the importance of gyres in the distribution ofCHP, which play an important role in the intensification of cyclones/depressions.CHP is >14 kcal/cm2 over Andaman Sea, southern and Central Bay of Bengal where the generation and movement of cyclones take place during post south west monsoon season (October–November). A depression formed on 07.11.95 at 11°N; 91°E and intensified into a cyclonic storm by 8th November evening and crossed Orissa Coast on 9th November 1995. A few days before its formation, the value of CHP at the origin of thiscyclone was about 20 kcal/cm2. To understand the exact role of CHP in theformation and intensification of cyclones/depressions over Bay of Bengal, more intense and systematic data sets are essential.

The many forms of bipolar disorder: a modern look at an old illness

Bipolar disorder continues to be underrecognized, despite being known for 2000 years. Mania, the fullest expression of the disease affects approximately 1% of the population; the less-than-manic forms of the disease dominated by depressive episodes have recently been found to be more common, affecting 4–5% of the population. In reviewing the international literature on this broadened bipolar spectrum, this paper pays particular tribute to the French EPIDEP and EPIMAN studies and Italo-American collaboration which have generated the largest set of systematic data on the new clinical portrait of bipolar disorders. Early detection is crucial, because untreated bipolar disorder has a high mortality rate. A review of the diagnostic criteria for the various subtypes of bipolar disorder has identified several factors that interfere with making an accurate diagnosis. These include age at onset, ethnic differences, co-morbidity (particularly substance abuse and alcoholism), and the broad range of clinical presentations. Moreover, symptoms frequently overlap with those of other psychiatric disorders including schizophrenia, attention-deficit disorder and personality disorders. Misdiagnosis is a major factor leading to a poor outcome for patients. Accurate identification and diagnosis of the different forms of mania can lead to specific treatment choices that may improve prognosis. Particularly important are recent data indicating reduced mortality with a variety of psychopharmacologic agents including, but not limited to, lithium and valproate.

The Lynching of Persons of Mexican Origin or Descent in the United States, 1848 to 1928

The lynching of persons of Mexican origin or descent has been largely over-looked by historians of American mob violence. This essay offers the first attempt to construct a systematic set of data on the subject. The authors contend that between 1848 and 1928, mobs lynched at least 597 Mexicans. Traditional interpretations of western violence cannot account for this phenomenon. The actual causes of mob violence against Mexicans were several-fold: race and the legacy of Anglo American expansion, economic competition, and diplomatic tensions between Mexico and the United States. Throughout this era, Mexicans formulated numerous means of resistance against Anglo mobs. These included armed self-defense, public protest, the establishment of mutual defense organizations, and appeals for aid to the Mexican government. The central aim of this essay is to broaden the scholarly discourse on lynching by moving beyond the traditional limitations of the black/white paradigm. Placing the experience of Mexicans into the history of lynching expands our understanding of the causes of mob violence and the ways in which individuals and groups sought to resist lynching and vigilantism. The essay is based on numerous archival sources in both Spanish and English. These include diaries, letters, memoirs, folk culture, newspapers, government documents, and diplomatic correspondence.

Variation of mechanical properties along the length of the aorta in C57bl/6 mice.

The objective of the present study is to obtain a systematic set of data on the mechanical properties along the entire length of the mouse aorta. The ascending aorta of seven mice was cannulated near the aortic valve, and the aorta was preconditioned with several cyclic changes in pressure. The perfusion pressure was then increased in 30-mmHg increments from 0 to 150 mmHg. Cab-O-Sil, colloidal silica, was mixed into the perfusate to prevent flow through the microvessels and hence attain zero-flow distensions. Our results show that the residual circumferential strain leads to a uniformity of transmural strain of the aorta in the loaded state along the entire length of the aorta. This uniformity is attained in the range of 60-120 mmHg. At pressures <60 mmHg, the outer strain is greater than the inner strain, whereas at pressures >120 mmHg, the converse is true. Furthermore, we found that the circumferential and longitudinal stress-strain relationships are linear in the pressure range of 30-120 mmHg. Finally, the circumferential modulus is greatest (most rigid) near the diaphragm, and the majority of volume compliance (85%) is in the thoracic compared with the abdominal aorta. These findings are important for an understanding of the hemodynamics of the cardiovascular system of the normal mouse and will serve as a reference state for the study of various diseases in knock-in and knock-out models of this species.

Shotgun crystallization strategy for structural genomics: an optimized two-tiered crystallization screen against the Thermotoga maritima proteome.

As the field of structural genomics continues to grow and new technologies are developed, novel strategies are needed to efficiently crystallize large numbers of protein targets, thus increasing output, not just throughput [Chayen & Saridakis (2002). Acta Cryst. D58, 921-927]. One strategy, developed for the high-throughput structure determination of the Thermotoga maritima proteome, is to quickly determine which proteins have a propensity for crystal formation followed by focused SeMet-incorporated protein crystallization attempts. This experimental effort has resulted in over 320 000 individual crystallization experiments. As such, it has provided one of the most extensive systematic data sets of commonly used crystallization conditions against a wide range of proteins to date. Analysis of this data shows that many of the original screening conditions are redundant, as all of the T. maritima proteins that crystallize readily could be identified using just 23% of the original conditions. It also shows that proteins that contain selenomethionine and are more extensively purified often crystallize in distinctly different conditions from those of their native less pure counterparts. Most importantly, it shows that the two-tiered strategy employed here is extremely successful for predicting which proteins will readily crystallize, as greater than 99% of the proteins identified as having a propensity to crystallize under non-optimal native conditions did so again as selenomethionine derivatives during the focused crystallization trials. This crystallization strategy can be adopted for both large-scale genomics programs and individual protein studies with multiple constructs and has the potential to significantly accelerate future crystallographic efforts.