Gozalo P.,Brown University |
Leland N.E.,University of Southern California |
Christian T.J.,AAI |
Mor V.,Brown University |
And 2 more authors.
Journal of the American Geriatrics Society | Year: 2015
Objectives To examine the effect of the relationship between volume (number of hip fracture admissions during the 12 months before participant's fracture) and other facility characteristics on outcomes. Design Prospective observational study. Setting U.S. skilled nursing facilities (SNFs) admitting individuals discharged from the hospital after treatment for hip fracture between 2000 and 2007 (N = 15,439). Participants Community-dwelling fee-for-service Medi-care beneficiaries aged 75 and older admitted to U.S. hospitals for their first hip fracture and discharged to a SNF for postacute care from 2000 to 2007 (N = 512,967). Measurements Successful discharge from SNF to community, defined as returning to the community within 30 days of hospital discharge to the SNF and remaining in the community without being institutionalized for at least 30 days, was examined using Medicare administrative data, propensity score matching, and instrumental variables. Results The overall rate of successful discharge to the community was 31%. Of the 15,439 facilities, the facility interquartile range varied from 0% (25th percentile) to 42% (75th percentile). An important determinant of variation in discharge rate was SNF volume of hip fracture admissions. Unadjusted successful discharge from SNF to community was 43.7% in high-volume facilities (>24 admissions/year), versus 18.8% in low-volume facilities (1-6 admissions/year). This facility volume effect persisted after adjusting for participant and facility characteristics associated with outcomes (e.g., adjusted odds ratio = 2.06, 95% confidence interval = 1.91-2.21 for volume of 25 vs 3 admissions per year). Conclusion In community-dwelling persons with their first hip fracture, successful return to the community varies substantially according to SNF provider volume and staffing characteristics. © 2015, Copyright the Authors Journal compilation © 2015, The American Geriatrics Society. Source
News Article | August 30, 2016
In this edition of the clean energy and transport news roundup, we’ve got stories about Colorado solar fees, sheep as solar workers, solar microgrids, a new integrated energy storage system, driverless car insurance policies, a 2500 mile EV roadtrip, automatic braking systems, Musk’s pessimism about slowing climate change, Chinese EV startups, and more. [CleanTechnica isn’t the only Important Media site to cover clean energy & transport news, and if you’re looking for more stories on electric mobility, bicycles, solar and wind energy, and other related issues, we’ve got them at sites such as Solar Love, CleanTechies, Planetsave, Bikocity, Gas2, and EV Obsession. We also host a large cleantech group on LinkedIn, called CleanTechies Around The World.] In Colorado, XCel Energy proposed a new charge on all solar customers, an idea that was furiously resisted by a dozen or more solar power advocacy groups. Now, all parties have agreed to a negotiated settlement of the dispute that may serve as a model for other states. The agreement involves refocusing the pricing of electricity for retail customers on a time of day model. Not every solar energy installation goes on top of a roof. Utility scale systems, community solar, and some residential systems often are mounted on the ground rather than up in the air. What many of the owners of such systems tend to forget is that steps need to be taken to control the vegetation that will inevitably start to grow up underneath and around the panels as soon as the project is completed. The plan was to install 200 solar panels, seven Tesla Powerwalls, several SolarEdge inverters, along with installing lights and plugs for dozens of buildings. In order to complete the project on time, they hired and trained local electricians and construction crews. After three weeks in Virunga, all three solar micro grids were successfully installed. Now the park rangers have reliable lighting, communication systems, computers, radios, and GPS units. That means they no longer need to risk their lives to bring in diesel fuel and can more effectively protect the park and its denizens from thieves and poachers. SimpliPhi Power technology utilizes patented, licensed lithium ferrous phosphate cells with state-of-the-art proprietary management boards, circuitry, cell architecture and methods of assembly to create safe, intelligen,t and energy dense storage and management systems. SimpliPhi batteries feature the highest efficiency rate in the industry (98 percent), 10,000 cycle life and 100 percent depth of discharge. The compact, lightweight form factor does not require ventilation or cooling, is rated for outdoor installation, and can operate at a wide temperature range of -4 to 140 degrees Fahrenheit. Alyssa Craft and Jesse Stafford completed a cheap off-grid cabin-add-on to the trailer they are living in while setting up their homestead in the Western US. The cabin was built to allow them to heat the air around the trailer and prevent the plumbing, and themselves, from freezing during the snowy winter months. Sri Lanka looks to 70% renewables by 2030: The ‘Battle for Solar energy’ program aims to encourage the small consumers to install solar panels at their roof-tops and consumers will be paid for any excess energy exported to the grid. With this new program, government expects that at least 20% of the consumers produce electricity on their own. India wind energy policy has developers apprehensive about country’s first wind energy auction: Project developers are apprehensive about India’s latest wind energy policy which calls for large-scale auction of projects. The uncertainty amongst wind energy project developers in India originates from the fact that there is lack of clarity on the extension of crucial incentives like the accelerated depreciation and generation-based incentive. These two support mechanisms have been the sole driver of wind energy expansion in India. Indonesia could see 19% returns on solar Feed-in Tariffs: Indonesia may soon become a budding solar power market as feed-in tariffs recently announced by the government are expected to yield very attractive returns for developers. The Minister for Energy in the southern state of Karnataka recently told media outlets that construction on a proposed 5 GW solar power park will begin soon. The Pavagada solar power park is among the more than two dozen solar power parks with cumulative capacity of almost 20 GW approved by the Ministry of New & Renewable Energy. According to media reports, the Airports Authority of India (AAI) has installed solar power panels on the cargo complex of the Chennai airport. The solar power system is believed to have installed capacity of 5 MW and will generate enough electricity to meet 80% of the cargo terminal’s power demand. Engie subsidiaries will develop 400 MW of solar in Chile: Engie Energia Chile and Solairedirect, subsidiaries of French utility Engie, have announced plans to jointly develop 400 MW of solar power in Chile. South Africa could have 5.6 GW of wind capacity by 2020: The research and consulting firm GlobalData report states that more than 3 GW wind energy is expected to be installed by 2020 which will bring country’s cumulative wind capacity to 5.6 GW. The company apparently intends for the launch of the new insurance policy to move forward “the discussion around who or what is liable in the event of a collision.” The policy was designed “for people who may have driverless or autonomous features in their existing car, or who may be thinking of buying a new car with driverless or autopilot features such as Tesla’s forthcoming Model 3.” The trip — which has been dubbed Route 57 — involved 2,500 miles of travel through England, Wales, Scotland, and Ireland; through various scenic towns, national parks, cities, etc. The trip was blogged about in great detail by the driver, Jess Shanahan. Orange EV taking orders for its new T-Series all-electric terminal truck: The new T-Series electric terminal trucks build on the company’s initial T-Series trucks (capable of up to 24 hours of operation per charge), which have now been deployed at various sites in the manufacturing, shipping, retail distribution, warehouse, railroad inter-modal, and waste management industries. And buyers could qualify for up to $120,000 in incentives: Orange EV has announced that organizations and businesses operating fleets in California will now be able to receive up to $120,000 in incentives towards the purchase of new T-Series all-electric Class 8 terminal trucks with extended-duty battery packs. Tom Moloughney is an avid electric car advocate. Two years ago, he purchased a BMW i3 with the 0.65 liter onboard two cylinder range extender gasoline engine. 27 months later, he has 56,000 miles on the odometer. He says 96% of those miles have been done on electric power alone. He has purchased about 50 gallons of gasoline for the car, which works out to be more than 1,000 miles per gallon. Mercedes will market its EVs under the EQ brand: No self respecting premium auto company would dive into the deep end of the electric car pool without dreaming up some snazzy new title to define its EV brand. BMW started it with its “i” division. Audi followed along with its E-Tron models. Now Mercedes has all but confirmed that it will market its electric cars under the EQ brand. 90% of Chinese EV startups could be wiped out: Soon, entrepreneurs will need a license from the government to become an automobile manufacturer. (Existing companies like SAIC and BYD will not need to apply for licenses.) According to reports, no more than 10 licenses will be issued — a very small number for such a large country. “There are too many entrants in the sector, and some of them are just speculators,” said Yin Chengliang, a professor at Shanghai Jiao Tong University’s Institute of Automotive Engineering. “The government has to raise the threshold. It’s bad to see irrational investments in projects with low technology levels.” The summer doldrums are almost over and the world of automobiles is about to get re-energized as the new auto show season kicks off in Paris in September. Citroen has gotten the buzz started early by showing off its latest creation, the CXperience large sedan concept. Sadly, it looks at though the Citroen design team has been heavily influenced by the new Toyota Prius. That is not meant as a compliment. Speaking with MSNBC host Chris Hayes recently. Elon Musk said he is pessimistic about slowing or reversing the effects of carbon pollution on global climate change. Doing so will require an effective means of pricing carbon emissions agreed to and implemented by all nations, he told Hayes. In his remarks, Musk made reference to what economists call The Tragedy of the Commons. There are two kinds of automatic braking systems — one designed to avoid forward collisions and one that only tries to minimize the damage from one. Which is which? How do you know which system your car is fitted with? And what are you, the driver, supposed to do — or not do — if the car ahead stops short suddenly or a pedestrian steps into your path? Drive an electric car? Complete one of our short surveys for our next electric car report. Keep up to date with all the hottest cleantech news by subscribing to our (free) cleantech newsletter, or keep an eye on sector-specific news by getting our (also free) solar energy newsletter, electric vehicle newsletter, or wind energy newsletter.
All the sequence data and metadata from the samples used in this work could be accessed through the Integrated Microbial Genomes with Microbiomes system IMG/M database11 (https://img.jgi.doe.gov) using both metagenome and scaffold identifiers provided throughout the manuscript and the Supplementary Information. Thus, by using these identifiers in the Genome Search tool or Scaffold Search tool (under ‘Find Genomes’ tab) in the user interface, the corresponding sequences, their annotations, as well as their associated metadata can be retrieved. Moreover, Hidden Markov Models (HMMs) of viral protein families as well as the predicted DNA viral sequences in fasta format are available at the following public FTP site: (http://portal.nersc.gov/dna/microbial/prokpubs/EarthVirome_DP/). All publicly available metagenomic data sets from the IMG/M system (3,042 samples comprising 5 terabase pairs of sequences) were used for this analysis11. The sample collection included 1,729 environmental samples, 1,079 host-associated samples, and 234 engineered samples according to the Genomes OnLine sample classification5. We identified putative viral contigs in 1,882 out of the 3,042 metagenomic data sets. The metadata for these data sets including sample collection information, library construction and sequencing protocols, as well as assembly strategy were retrieved from GOLD database5. Based on GOLD metadata, the vast majority of these data sets were generated from dsDNA using an untargeted approach (that is, only 59 samples underwent viral particle enrichment, viral DNA enrichment or library construction with sequencing protocols optimized for the recovery of viral sequences). All of the data sets have been annotated by the IMG metagenome annotation pipeline44, which performs gene prediction and functional annotation through assignment of predicted proteins to protein families, such as Pfam45 and KEGG Orthology (KO) clusters46. Some of the data sets included both assembled and unassembled data, while others had only assembled sequences (Supplementary Table 8). An assembly pipeline used for each data set is described in GOLD. In addition, the contiguity of assembled sequences varied greatly from sample to sample. The ecosystem subcategories here used were manually curated according to sample metadata establishing 10 distinct habitat types: marine, freshwater, non-marine-saline and alkaline, thermal springs, terrestrial soil, terrestrial others (including mostly deep subsurface samples), host-associated human, host-associated plants, host-associated others (including host animal-associated other than human), and engineered (for example, bioreactor) (Supplementary Table 8). Only contigs longer than 5 kb (59.5 Gb from 5.1 million contigs) were primarily included in this study. We normalized the data sets by the size of the sample (measured as total number of bp from sequences larger than 5 kb) per habitat type. The normalization factor used in each habitat type was: marine, 29,602 Mb; freshwater, 96,314 Mb; non-marine saline and alkaline, 2,825 Mb; thermal springs, 1,828 Mb; terrestrial (soil), 5,794 Mb; terrestrial (other), 1,659 Mb; host-associated (human), 10,349 Mb; host-associated (plants), 3,909 Mb; host-associated (others), 23,452 Mb; engineered, 10,486 Mb. We used a combination of 2,353 iVGs composed of all isolate dsDNA viruses and retroviruses from the NCBI server (http://www.ncbi.nlm.nih.gov/genome/viruses/, data accessed on 04/2015) to extract all viral proteins and to establish, after filtering, the first round of viral protein families. Additionally, we used a list of 5,042 reference viruses (Supplementary Table 13) extracted from the IMG/M system (that included all RNA and DNA eukaryotic and prokaryotic referenced viral genomes) to generate and validate our viral genome clustering method and also to calculate the average length of all reference viruses as 44,296 bp ± 83,777 bp s.d. (Supplementary Table 13). 167,042 protein coding genes were collected from 2,353 iVGs (dsDNA viruses and retroviruses combined) from the NCBI server (http://www.ncbi.nlm.nih.gov/genomes/GenomesGroup.cgi?taxid=10239#). After dereplication using 70% identity in usearch47, 98,000 protein sequences were obtained from which 83,500 were clustered into 15,900 groups using the Markov Cluster (MCL) algorithm48. Proteins within clusters were aligned using MAFFT49 and a set of 14,296 viral protein families was created using hmmbuild50. After manual curation of the viral families with high representation in prokaryotic genomes, viral protein families were compared against the 5.1 million metagenomic contigs longer than 5 kb. 62,000 contigs with 5 or more viral protein families were collected, and these were reduced to 9,000 putative viral contigs after removing contigs below 50 kb. An additional filtering step was performed to exclude contigs with a high number of Kegg Orthology (KO) terms and Pfams (10% and 25% respectively); this reduced the number of putative viral contigs to 1,589. These were complemented with 66 and 188 sequences derived from diverse metagenomic contigs longer than 20 kb that were binned with viruses or contained a viral RNA polymerase gene, respectively, and were not captured using the previous filter of bearing 5 or more viral protein families (detailed in section below; Extended Data Fig. 2; Supplementary Table 1). A total of 1,843 mVCs encoding 191,000 proteins were used to complement the original set of 167,042 proteins derived from iVGs. Repeating the steps described above (that is, usearch 70% for de-replication, MCL clustering, MAFFT alignment and hmmbuild with a filter for viral families abundant in prokaryotes) the final list of 25,000 viral protein families was obtained and used for further exploration. To expand the training set of viral sequences, metagenome contigs identified as high-confidence viral sequences in the first iteration of our pipeline (Extended Data Fig. 1) were complemented with additional metagenome contigs and scaffolds, not captured using viral protein families generated from isolate viruses. The first approach used kmer-based binning of 6 metagenome samples that contained the highest number of candidate viral sequences, which were not satisfy high-confidence threshold due to insufficient number of hits to protein models. These data sets were binned by Emergent Self Organizing Maps (ESOM; by Ultsch) as described previously51 and contig sets outside the bins corresponding to cellular organisms were manually checked (Extended Data Fig. 2a). K-mer-based binning identified 66 putative novel mVCs from diverse habitat types (freshwater, wastewater, thermal vents, and marine with IMG sample identifiers 3300000553, 3300001592, 3300001681, 3300000116, and 3300001450, respectively). The second approach relied on identification of contigs containing RNA polymerase with domain composition reminiscent of RNA polymerase (RNAp) found in cellular life forms, which could not be placed into one of three domains on the tree of life based on their sequence similarity. First, 2,551 representative sequences of the genes encoding the three major subunits (α, β, β’) of the RNAp gene from bacteria, as well as their eukaryotic and archaeal counterparts, were collected from IMG database. Next, the domains of these genes were extracted using Pfam models and aligned with MAFFT49. Alignments were manually inspected and HMM models were built using hmmbuild50. These models were used to scan metagenomic sequences longer than 5 kb and identified 39,109 contigs with matches for at least one core RNAp subunit. After filtering short matches and a dereplication step, we obtained 7,437 metagenomic sequences that were combined with 2,551 reference isolates to build a tree with 9,309 RNAp sequences using FastTree52 with default parameters (Extended Data Fig. 2d). The tree was visualized using Dendroscope53 and RNAp branch corresponding to large eukaryotic DNA viruses was identified on the basis of reference sequences from isolate genomes. In addition to eukaryotic viruses, another set of metagenomic RNAp sequences branching separately from cellular references, turned out to comprise phage RNAp with domain composition similar to bacterial enzyme (detailed in Extended Data Fig. 2e). A total of 188 contigs longer than 20 kb containing viral and phage RNAp sequences were added to the training set. The 25,000 viral protein families were used to identify 125,842 DNA metagenomic viral contigs (mVCs) longer than 5 kb using 3 distinct filters. First, mVCs were identified from metagenomic contigs that had at least 5 hits to viral protein families, the total number of genes covered with KO terms on the contig was <20%; the total number of genes covered with Pfams ≤40%; and the number of genes covered with viral protein families >10%. Second, metagenomic sequences were selected as mVCs when the number of viral protein families on the contig were equal or higher than the number of Pfams. Finally, metagenomic contigs for which the number of viral protein families was equal or higher than 60% of the total of the genes were also assigned to mVCs. Benchmarking and modelling of this DNA viral discovery computational approach are detailed below, demonstrating a specificity of 99.6% for viral detection with a 37.5% recall rate (sensitivity to identify all viral sequences). In order to assess the accuracy of our DNA vHMM virus detection pipeline, we generated a synthetic metagenome, consisting of finished genomes of 32 bacteria, 3 archaea and 5 viruses (Supplementary Table 2), which included 88 replicons. Bacterial genomes include representatives of 4 phyla. A total of 132 prophage sequences were identified including 99 prophages identified by CyVerse54 implementation of VirSorter55 in the categories 1, 2, 4, and 5, and 33 prophages identified by manual curation based on the presence of hallmark phage genes and analysis of synteny with closely related strains. Coordinates of 35 prophages predicted by VirSorter had to be manually adjusted to eliminate bacterial genes (including ribosomal RNAs and other housekeeping genes) and to separate 2 prophage sequences called as one prophage over an intervening stretch of bacterial genes. Coordinates of the prophages are provided in Supplementary Table 3. None of the viruses or prophages used in the synthetic metagenome were included in the training set used to generate viral HMMs for our pipeline. The genome sequences were fragmented to generate 63,222 contigs of length 5 kb to 60 kb. The distribution of fragmented contigs include 28,497 5-kb-long fragments, 14,228 10-kb-long fragments, 7,096 20-kb-long fragments, 4,723 30-kb-long fragments, 3,525 40-kb-long fragments, 2,810 50-kb-long fragments and 2,343 60-kb-long fragments. The resulting synthetic metagenome is dominated by bacterial and archaeal chromosomal fragments with an admixture of a relatively small number of plasmid and viral sequences, which is a faithful representation of a typical metagenome data set generated by an untargeted approach rather than by targeted virome sequencing approach. The metagenome was submitted to a CyVerse implementation of VirSorter and also processed by our vHMM pipeline. Only the categories 1, 2, 4 and 5 of Virsorter predictions were considered, as manual inspection showed that categories 3 and 6 contained mostly false positives. Sequence fragments with at least 3 kb of phage or prophage sequence were considered as true positive viral sequences; those with less than 3 kb of phage or prophage sequence were considered true negative. The 125,842 metagenomic viral contigs longer than 5 kb encoded a total of 2.79 million proteins. BLASTp56 with an e-value of 1.0 × 10−5 was used and 1 hit per query protein with >60% sequence identity and >80% alignment on the shorter sequence. Proteins encoded by mVCs were clustered using CD-HIT57 at 60% sequence identity and >80% alignment on the shorter sequence. For each sample count, 100 random metagenome sets were generated and the total number of protein clusters found on the contigs from this set was calculated. This analysis was repeated separately for metagenome samples classified as ‘aquatic’ (n = 656) and ‘human’ (n = 673). Sequence similarity of mVCs to iVGs was computed using BLASTp56 with an e-value threshold of 1.0 × 10−5 and alignment length of at least 80% of the shorter protein. No percentage identity or bit-score cutoffs were applied (Supplementary Table 9). To assess the number of closed DNA mVCs, we searched for overlapping sequences in the 3′ and 5′ region of all the 125,842 metagenomic contigs. Extractseq58 was used to trim the first 100 bp of each contig and BLAT59 was used to search each 100-bp fragment against the respective contig. Only exact overlapping matches for both the 3′ and 5′ regions were considered. This resulted in the identification of 999 putatively closed mVCs, ranging from 5,037 bp to 630,638 bp in length (average, 53,644 bp ± 45,677 bp s.d.). Supplementary Table 10 lists all putatively closed mVCs. A sequence-based classification framework was developed for systematically linking closely related viral genomes based on their overall protein similarity. The framework relies on both AAI and total alignment fraction (AF) for pairwise comparisons of viral sequences, and enables natural grouping of related iVGs and mVCs. The 125,842 mVCs were combined with all iVGs (DNA and RNA viruses) for the generation of the viral group classification framework (Supplementary Information). To reduce the number of the AAI comparisons, only mVCs that contained at least one protein match with ≥70% identity across ≥50% of the shortest protein length were selected for pairwise computations. This filter reduced the number of total pairwise comparisons from 9.5 billion to 15.9 million. The bidirectional average amino acid identity (AAI) was performed as previously described18 for all of the 15.9 million pairwise comparisons. This method implements usearch47 for rapid blast, and selects the bidirectional best hit for each protein encoded on the mVC and outputs the AAI and the AF. The output was subsequently filtered to include only matches that had ≥90% AAI and ≥50% AF which were the observed parameters that best reproduced the existing taxonomy of iVGs (Supplementary Information; Extended Data Fig. 5a). The high-quality filtered AAI results were then clustered using single-linkage hierarchical clustering and visualized in Cytoscape60 (Extended Data Fig. 5c–e). As a validation of our clustering method we observed that 87% of the iVGs (920 out of the 1,060 viral groups or singletons) with a taxonomic assignment according to the International Committee on Taxonomy of Viruses (ICTV) clustered in agreement with the ICTV-designated species. All the remaining 13% of iVGs clustered at the genus-level. From this 13% (represented by 140 viral groups that contain at least one iVG) we found that only 49 were phage groups, with high pairwise (over 90% AAI) values for the reference viruses within each group, suggesting that despite their taxonomic assignments, they were also probably members of the same species (Supplementary Table 14). These analyses show that our viral groups are taxonomically relevant and provide a useful method for organizing distinct viral types. A CRISPR–Cas spacer database of 3.5 million sequences was created using a modified version of the CRISPR Recognition Tool61 (CRT) detailed in ref. 44 against 40,623 isolates and 6,714 metagenomes (all data sets from the IMG system as of 9 July 2015). All identified spacers were queried for exact sequence matches against all iVGs using the BLASTn-short function from the BLAST+ package with parameters: e-value threshold of 1.0 × 10−10, percentage identity of 95%, and using 1 as a maximum target sequence56. 98.5% of the detected 1,340 spacer hits were to a putative bacterial or archaeal host whose taxonomic assignment was in agreement at the species or genus level with the existing viral taxonomy (Supplementary Table 18). From the remaining matches, 1.2% of the hits agreed at the family level and only 0.3% of the spacers (2 cases where Pseudomonas spacers matched a Rhodothermus phage, and Methylomicrobium spacers that matched Pseudomonas and Burkholderia phage) were above family, validating our approach of host assignment based on CRISPR–Cas spacer matches. Subsequently, all 3.5 million spacers were compared against the 125,842 mVCs, requiring at least 95% identity over the whole spacer length, and allowing only 1–2 SNPs at the 5′ end of the sequence. A total of 12,576 proto-spacers (that is, spacer sequences within a phage genome) were identified. Based on CRISPR–Cas spacer matches exclusively from microbial isolate genomes we assigned host taxonomy to 8,084 mVCs (representing 6.42% of all the mVCs), comprising 826 viral groups (~4.47% of the total) plus 1,100 viral singletons (~1.71%) (Fig. 2a; Supplementary Table 19). Identification of tRNAs from mVCs was performed with ARAGORN v1.2 (ref. 62) using the ‘-t’ option. In order to validate this approach, 2,181 tRNA sequences were recovered from 344 referenced viruses (~7% of the total). These were compared against all genomes and metagenomes in the IMG system using BLAST, leading to 16,089 perfect hits (100% length and 100% sequence identity) after removing self-hits and duplicates. The taxonomic assignment of the tRNAs found in iVGs was compared against the taxonomic information of the isolate microbial genomes showing that 92.5% of the matches agreed at the genus or species level (Supplementary Table 18). After culling the top-20 most abundant viral-tRNA sequences (sequences conserved across members of the gammaproteobacteria class; Supplementary Table 22) and repeating the above steps with mVCs, 32,449 tRNAs within 9,555 mVCs (7.6% out of the 125,842 total) were identified, enabling the host assignment for 2,527 mVCs (Supplementary Information; Supplementary Table 19). In order to detect the presence of any of the mVCs in lower abundances across different habitat types, we expanded our analysis to include not only assembled data (that probably represent the most abundant viruses) but also unassembled data from 4,169 samples currently available in IMG/M database, which comprises more than 5 Tb of sequences. We used BLASTn program in the Blast+ package56 to find hits to our 125,842 predicted viral sequences with an e-value cutoff of 1 × 10−50, at least 90% identity, and the hits from the sample covering at least 10% of the length of the viral contig. This filtering of BLAST results excluded matches to short highly conserved fragments of viral sequences, such as tRNAs, and other spurious hits. Our filtering criteria were optimized for the type of metagenome data sets available to us, and are significantly more stringent than those used in some previous studies for similar data (e.g. 95% identity over 75 nt alignment used in ref. 63) or tBLASTx with e-value of 1.0 × 10−5 recommended by ref. 64. However, it was less stringent than the 75% coverage used in the analysis of Tara Oceans Viromes3, which relied on viral enrichment to increase viral sequence coverage. For the largest metagenome available to us (IMG taxon 3300002568, Grasslands soil microbial communities from Hopland, California, USA), this new analysis was able to detect 500 nt of viral sequence in 138,769,704,035 nt of total metagenome sequence, which corresponds to the abundance of 3.06 × 10−07%. Habitat type specificity of predicted viral sequences based on their BLASTn hits in assembled and unassembled data shows their presence even at low abundance, depending on the sequence coverage for each specific metagenome (Fig. 4a, b). The distribution of less abundant viruses supports the trend that viruses have a strong specificity for a particular habitat type since ~84% of all the viral groups are found exclusively in a single habitat type. About 14% of the viral groups were found in 2 habitat types, and most of these cases could be explained by the uncertainty of habitat type classification. For instance, algae-associated microbiomes were classified as plant host-associated and shared viral groups with marine samples, whereas loose soil samples classified as terrestrial habitat type shared viral groups with rhizosphere samples, which were classified as plant host-associated. After excluding ambiguously classified cases, most viral groups detected in more than 1 habitat type were found in the samples from the same environmental category (for example, in different aquatic habitats or in different mammalian hosts). We further report the finding of ~0.2% of the viral groups in 5 or more habitats types and discuss the main types of these ‘cosmopolitan’ viruses (probably laboratory contaminants, prophages with broad-host specificity, and bona fide lytic phages with unexpectedly broad habitat type distribution). Raw reads for faecal and oral metagenomes were retrieved from the Short Read Archive (http://www.ncbi.nlm.nih.gov/sra/) based on the metadata available in GOLD. The reads were quality-filtered and quality-trimmed using rqcfilter tool from BBtools package (https://sourceforge.net/projects/bbtools/) with default settings: kmer length for trimming of 23, minimum average quality of 5, trim quality threshold of 10, reads shorter than 45 nt after trimming were discarded. Quality-filtered and trimmed reads were digitally normalized and error corrected using bbnorm tool from BBtools package with default settings. Normalized reads were assembled using SPADES 3.6.2 (ref. 65) and kmers of 19, 39, 59, 79, 99, selecting an optimal kmer length based on the maximal N50. Average contig and scaffold coverage of assembled data was calculated by mapping the quality-filtered and -trimmed reads to the assembly using bbmap tool from BBtools with default kmer length of 13 and minimum percentage identity cutoff of 95%. The unmapped reads were merged using bbmerge tool from BBtools package and the sequences shorter than 100 nt were discarded. mVCs were aligned against these data using BLASTn and filtered as described above. Only 1 best hit per sequence was retained. Coverage of each mVC by sample data was calculated as alignment length multiplied by the coverage of the subject sequence and summed over all sequences in the sample with hits to this mVC. We have identified putative prophages among 125,842 mVCs using these contigs as a query and running BLASTn56 comparison of ‘blast+’ package against all isolate genomes in the IMG database. e-value cutoff of 1.0 × 10−50 and percentage identity of 80% were used, and mVCs with cumulative alignment of at least 75% of mVC length against an isolate genome were considered prophage candidates (Supplementary Table 4). Visualization was made with the use of Processing programming language (https://processing.org/) and a freely available equirectangular projection of the world map (http://eoimages.gsfc.nasa.gov/images/imagerecords/57000/57752/land_shallow_topo_2048.jpg) was used as a background image. Sample points are positioned by latitude and longitude coordinates of Biosamples obtained from GOLD5. Points are coloured based on a customized reclassification of the GOLD hierarchical ecosystem classification (habitat types). Lines between points indicate samples that share at least 2 viral groups or singletons.
Simon S.,AAI |
Archives of Phytopathology and Plant Protection | Year: 2011
Onion is an important indispensable item in every kitchen as condiment and vegetable, hence commands an extensive internal market. The green leaves and immature and mature bulbs are eaten raw or used in preparation of vegetables. Among the various pests and diseases associated with onion, the root knot nematode (Meloidogyne incognita) had proved itself as an important limiting factor for successful cultivation of this crop. Recently root-knot nematode was observed to cause serious losses to green onion crop grown mostly in the field located near the Jamuna & Ganga river belt of Allahabad and the surrounding area in U.P. Differential host test was done to confirm the root knot disease in onion. The specific identity of the nematode was determined by cutting perineal pattern of the females and was confirmed as Meloidogyne incognita. © 2011 Taylor & Francis. Source
Archives of Phytopathology and Plant Protection | Year: 2011
Dactylaria eudermata Drechsler is one of the important predaceous fungi occurring widely in different soils. The fungus produces hyphal bails and network compound in which nematodes are entangled. This hyphal nets as trapping structure which may be single dimensional to two or three dimensional and are formed through repeated hyphal anastomosis, which capture nematodes either through the adhesive material present on the surface of hyphal nets or due to physical entanglement. In the experiment it was observed that inflation of hyphal bails takes 30 to 50 minutes and capturing and killing the nematode by a single conidia in water takes 35-55 hours, were as time required for inflation of hyphal bails and real trapping and killing of a nematode in maize meal (MMA: water (1:10) medium) takes one minute and 30-50 hours respectively. © 2011 Taylor & Francis. Source