English 日本語

What is TogoDX?

TogoDX (Togo Data eXplorer) is a framework for the integrated exploration of various databases that have been turned into knowledge graphs. It integrates many datasets derived from different databases, based on the compatibility of IDs that refer to the same entity and on ID relations that organize the connections between different kinds of data. TogoDX aims to provide a new mechanism for flexibly extracting data useful for data science, by narrowing down data with the diverse attributes that the datasets have.

Background

Problems

In the life sciences there are many databases dealing with different subjects. Processing them in an integrated manner and using them for data science involved the following problems.

Interfaces differ from database to database

Each database has its own interface. Because the way to access the website, the way to specify search conditions, and the terminology all differ, users have to learn how to use each database individually. The more databases are involved, the higher the cost of learning how to use them.

To combine data stored in multiple databases, meaningful correspondences (link information) have to be prepared. Even databases that do not list the IDs of external links may be connected by way of other databases. This requires users to build the link information themselves, and the more databases are involved, the higher the cost of building the links.

Output formats differ from database to database

The output formats used by databases vary. The work of converting data in different formats, derived from multiple databases, into a format that can be used in an integrated manner also becomes more costly as the number of databases increases.

Data integration with knowledge graphs and the development of TogoDX

To solve these problems, which had so far been left to database users, NBDC/DBCLS has been integrating life science databases with knowledge graphs. Using RDF (Resource Description Framework) as the technical foundation of the knowledge graphs brings the following advantages.

Data expressed in RDF can be searched and processed by writing queries in SPARQL, its query language. Integrating databases with RDF has made it possible to search and retrieve information across different datasets.

On the other hand, the integrated data has become very large and complex, and searching it requires a certain degree of expertise. TogoDX was therefore developed to process the integrated data at high speed and to make it explorable through operations in a web browser.

Note that TogoDX is a general-purpose framework: as long as the data is integrated with a knowledge graph, it can handle information from any domain, not only biology.

About TogoDX/Human

What is TogoDX/Human?

TogoDX/Human uses the TogoDX framework to provide an interface for the integrated exploration of data related to humans. It collects and integrates information on genes, proteins, compounds, diseases and more from databases in Japan and abroad. This site is developed and provided by NBDC/DBCLS.

Features of TogoDX/Human

TogoDX/Human lets you explore, in one place, information on humans that has so far been scattered across a wide variety of databases. Currently, 50 attributes (attributes extracted from each dataset) derived from 21 datasets are available.

How to use

TogoDX can be used in two ways: exploring data while getting an overview of all the data, and mapping your own data as IDs. It aims to be a framework whose results lead to further analysis.

Search conditions in TogoDX can be saved as a URL and shared with other users.

Update history

Recommended browsers

Funding

This service was developed as part of the Life Science Database Integration Project of NBDC.

Site policy