Using PZLAST, users can search the protein sequences to several Tera-bytes of public metagenomic amino acid sequences in approximately 10-20 minutes.
PZLAST uses PEZY-SC2, which is a MIMD-type many-core processor [1][2][3][4]. The algorithm of PZLAST is basically similar to CLAST [5].
References
PEZY-SC2 (in English)
[1] ZettaScaler: Liquid immersion cooling Manycore based Supercomputer
[2] pzqd: PEZY-SC2 Acceleration of Double-Double Precision Arithmetic Library for High-Precision BLAS
PEZY-SC2 (in Japanese)
[3] PEZY Computing: PEZY-SC2モジュール&プロセッサ
[4] グリーンスーパーコンピュータ ZettaScaler の技術と今後の展望
CLAST
[5] CLAST: CUDA implemented large-scale alignment search tool
Citation
If you find PZLAST useful in your research, please cite the following paper:
Hiroshi Mori, Hitoshi Ishikawa, Koichi Higashi, Yoshiaki Kato, Toshikazu Ebisuzaki, Ken Kurokawa, PZLAST: an ultra-fast amino acid sequence similarity search server against public metagenomes, Bioinformatics, 2021;, btab492, https://doi.org/10.1093/bioinformatics/btab492
- Contact
- This web site is operated by Genome Evolution Laboratory at the National Institute of Genetics and Computational Astrophysics Laboratory at RIKEN.
- If you have any question, request, or suggestion, please contact the E-mail address below.
- pzlastml.riken.jp
- v. 0.3.1 (2022-07-20)
- Added "Sample-wise results" to the results page.
- Completion-table, -MEO cloud, -world map results are organized into the "Sample-wise results" tab.
- v. 0.3.0 (2022-07-13)
- Added two features to the results page.
- One is environmental coverage information for each amino acid residue, and the other is a table of query hit fulfillment rates for each sample.
- v. 0.2.1 (2021-12-17)
- Reference database has been updated.
- As the number of reference sequences increased from 42 Billion to 64 Billion, the processing time of the search increased by about 20%.
- v. 0.2.0 (2021-8-20)
- Alignment information has been added to the output table.
- v. 0.1.0 (2020-4-1)
- PZLAST Web Service: First release.