dblast Software as a Service for rRNA-based Identification

dblast is our Software as a Service (SaaS) system for the GxP-compliant identification of Bacteria and Fungi based on ribosomal RNA gene sequences within regulated environments.

Your query sequences are uploaded to a login-protected online server and immediately analysed, any single analysis is done in less than 3 seconds. Your results can be viewed in the web interface of the system and downloaded as a pdf report for approval and archiving.

Every query sequence is searched against two comprehensive and high-quality reference databases of Bacteria and Fungi in parallel. The bacterial subset represents nearly all publicly available valid bacterial species (full length SSU rRNA). The fungal subset covers a representative subset of organisms relevant for identification in typical routine applications (D2 region of the LSU rRNA).

Main Features

• more than 14k bacterial and 2,3k fungal reference species, respectively
• up-to-date nomenclature and regularly updated -> identification represents current scientific status
• integrated user management, organize your complete team
• fully validated, used within GMP-regulated environments

Target environments

The dblast service was developed for laboratories which rely on highly reliable identification results (species level) based on ribosomal RNA gene sequence analysis and which operate within highly-regulated environments such as quality control laboratories in pharmaceutical industry.

The dblast reference database and reporting

The dblast reference database consist of two subsets of ribosomal rRNA gene sequences which are searched and reported in parallel for every of your query sequences:
A set of nearly all validly described bacterial species (full length 16S rRNA gene) and a set of selected fungal species of taxa relevant for diagnostics purposes (commonly used D2 region of the 28S rRNA gene).

Bacteria v11.0: 14.794 reference entries (species)

Fungi v2.7: 2.398 reference entries (species)

The dblast reference database follows the idea of the The All-Species Living Tree Project (LTP) of providing a single reference per validly described bacterial species, represented by only public type strain sequences.
This approach provides three major advantages:

(i) clear identification results and assessment (due to the absence of redundancy)

(ii) use of only original type strain sequences guarantees correctly assigned species names

(iii) starting from the complete public repositories results in a scientifically comprehensive and up-to-date database resource

However, compared to the academic LTP project (the highest quality SILVA-related subset), the dblast reference database has been subjected to intensive additional quality management which even meets industrial GxP-requirements. This is of major importance for professional routine applications because not all publicly available type strain sequences meet the standards required.

In conclusion, the dblast solution represents a direct implementation of our mission to make available the status quo of microbial systematics for industry-grade applications (see our Networking page).

Reporting:

For maximum overview and easy approval, the dblast system does not report its results as a list of always the best n hits, but instead uses the hit list organization scheme outlined below.
In this context, the system follows the concept of a 99% sequence identity threshold for reliable rRNA gene-based microbial species identification and a 97% cut-off for genus assignment.

• if the best hit shows 100% identity - all other hits of 100% are listed (if true) and the first hit <100%
• if the best hit shows <100% to 99% identity - all hits of this category are shown and the first hit <99%
• if the best hit shows <99% to 97% identity - all hits of this category are shown and the first hit <97%
• if the best hit shows <97% identity - the best 20 hits are shown

What is the meaning of 'dblast'?

The name dblast stands for 'diagnostic blast' and briefly describes how the tool works and what you can get out of it.

When users read 'BLAST', they usually think about the online service provided by NCBI which is mainly used to search the entire primary GenBank database, often resulting in unsatisfactory hit lists in terms of complexity and precision. Reason is the mass of redundant, incomplete and not quality-controlled information deposited by the community. But more correctly, BLAST (Basic Local Alignment Search Tool) is a generic toolbox for local installation based on a specific algorithm.

The original BLAST manuscript is one of the most-cited papers in scientific literature and since two decades the toolbox is considered to be an industry-standard workhorse. Therefore, we decided to apply selected BLAST components in the dblast core module for robust initial sequence alignment and search. However, the system does not provide a standard BLAST output to you. BLAST parameters have been optimized for rRNA gene reference databases and an advanced post-processing routine filters the BLAST output for potential invalid hits (based on partial alignments) and also categorizes the results.

In combination with the high-quality reference database (see above), the dblast system provides highly reliable species identification within the resolution capacity of the ribosomal RNA gene. This is why we decided for the extension 'diagnostic', to express the accurateness of the system.

Quality assurance

The dblast system was developed and is operated by us under the guidelines of Computerized System Validation (CSV) and we are running dedicated instances for customers from pharmaceutical industry and regulated by GMP/GAMP requirements (Good Manufacturing Practice/Good Automated Manufacturing Practice). In terms of quality assurance and rRNA gene-based identification capacity, there is no comparable bioinformatics solution on the market.