バージョン 27 (更新者: wu, 12 年 前)

--

Triple Store Survey for Life Science Data

Overview

We present an evaluation of native triple stores on biological data. Compared with the data in other areas biological data is typically huge. Therefore the performance of bulk loading and querying are essential to decide whether a triple store can be applied into the biological field. Here we test five native triple stores Virtuoso, OwlimSE, Mulgara, 4store, and Bigdata with five biological dataset, which ranging from tens of millions to eight billions. We present their load times and query cost.

For each database we provide several results by adjusting their parameters, which could influence the performance importantly but work differently with different hardware and software platforms. We do not guarantee what we provide is the best performance of each database.

4store

4store is a RDF/SPARQL store written in C and designed to run on UNIX-like systems, either single machines or networked clusters.

Bigdata

Bigdata is designed as a distributed database architecture running over clusters of 100s to 1000s of commodity machines, but also can run in a high-performance single-server mode. It supports RDFS and limited OWL inference. Bigdata is written in java and open source.

OwlimSE

OwlimSE is a member of OWLIM family, which provides native RDF engines implemented in Java and deliveries full performance through both Sesame and Jena. From OwlimSE 4.3 it begins to support SPARQL 1.1 Federation. OwlimSE is only available in commercial license.

Mulgara

Mulgara is written entirely in Java and available in open source. Mulgara provides a SQL-like language iTQL(Interactive Tucana Query Language) shell to query and update Mulgara databases. It also provides a SPARQL query parser and query .

Virtuoso

Virtuoso provides a triple storage solution for RDF in RDBMS platform. Virtuoso is multi-protocol RDBMS for RDBMS, RDF, XML and so on. It offers stored procedures to load RDFXML, ntriples, and compressed triples and supports for SPARQL. The function as a standalone triple store server is available in both open source and commercial licenses.

Platform

* Machine:

  • OS: GNU/linux
  • CPU: GenuineIntel? 6; model name : Intel(R) Xeon(R) CPU E5649 @ 2.53GHz; 12 cores 24 hyper-threading
  • Mem: 65996128 kB
  • Harddisk: SCSI Raid 0 (three hard disks of 2 Tera bytes; two of them are used to store data)

* Software:

  • JDK:1.6.0_26
  • 4store: 1.1.4
  • Bigdata: RWSTORE_1_1_0
  • Mulgara: 2.1.12
  • OwlimSE: 4.3.4238
  • Virtuoso: 6.4 commercial

Data

Cell cycle: .rdf (RDFXML) format, 11,315,866 tripples,from  http://www.semantic-systems-biology.org/. sparql query attachment:cell.txt ダウンロード .

Allie: .n3 format, 94,420,989 tripples, sparql query attachment:allie.txt ダウンロード .

PDBJ: .rdf.gz format ,589,987,335 triples, 77878 files, from  ftp://ftp.pdbj.org/XML/rdf/. sparql query attachment:pdbj.txt ダウンロード.

The queries in PDBJ are point queries which retrieve the relative characteristics of certain EntryID, such as 107L. Therefore their result set is small but the number of query joins is big.

Uniprot: .rdf.gz format , 4,025,881,829 triples, the 3 larger files are uniprot.rdf.gz,uniparc.rdf.gz,uniref.rdf.gz, from  ftp://ftp.uniprot.org/pub/databases/uniprot/ (the experiment used data was 2011.Nov version). sparql query attachment:uniprot.txt ダウンロード or  http://beta.sparql.uniprot.org/.

DDBJ: .rdf.gz format, 7,902,743,055 triples, 330 files, from  ftp://ftp.ddbj.nig.ac.jp/ddbj_database/ddbj/. sparql query attachment:ddbj.txt ダウンロード .

Approach

We evaluated the data in every Sparql end point at least twice to make it sure that there is no much difference between two test values:|2nd-1st|/max(2nd,1st)<0.1.

We did the query evaluation by executing the whole query mix (composed of the query sequence) five times in every Sparql endpoint, remove the highest one and then get the average time cost of other four queries. We report the five detailed time cost in every database section and the average cost in the summary section.

添付ファイル