March 15, 2011 (Vol. 31, No. 6)
IO Informatics Attempts to Make Life Science Research Easier with Suite of Software Products
Making sense of complex genomic and proteomic data is becoming easier as a result of advances in bioinformatics. Imagine connecting data in a fraction of the previously required time, and also incorporating information from published scientific papers, patient responses to medication, or complex histopathology images. That’s exactly the goal at IO Informatics. Its semantics integration software reportedly allows researchers to look at different types of data simultaneously, regardless of the source of the information or where it is located.
“Scientists have done a great job of creating data in specialized fields, but to make it useful, we have to meaningfully connect the dots across disciplines,” says Robert Stanley, CEO. Semantic technology offers a new way to link data quickly and flexibly across domains and give meaningful results to the end user. Stanley calls data integration a “persistent pain point” for the biological sciences, especially as data keeps expanding and changing. He and Erich Gombocz, Ph.D., CSO and vp, started IO Informatics in 2003.
The “IO” in the company’s name refers to intelligent multidimensional object (IMO) database records. IMOs reflect the underlying principle behind the Semantic Web, a consortium that helps computers to understand the meaning (or semantics) of information on the web. IMOs turn data into freeform relational objects that can be easily shared, manipulated, integrated, and compared. Stanley and Gombocz simplified the term to “intelligent object” when applied to their company’s intellectual property.
The company’s Sentient Suite™ platform was designed to analyze a range of data types. While focusing on genes, proteins, or compounds of interest, researchers can explore different databases. They can select a dataset, then drill down to a finer level and integrate it with other data. Scientists can also easily fit data from their own systems into published pathways, interactions, or other correlation networks. No matter where data is stored, it can be integrated and analyzed, Stanley reports.
Sentient Suite works with numeric values, images, spreadsheets, video, web content, public and private databases, or output from instruments like gene sequencers and mass spectrometers. Sentient Suite runs on laptops or personal computers and requires no further programming for integration. Basic relationships between data types can be viewed as networks, charts, dendrograms, matrix clusterings, tabular reports, or spreadsheets.
A key feature of Sentient Suite is the Knowledge Explorer. In December 2010, the company released version 3.3 of Knowledge Explorer, making it even easier to access and integrate data and uncover hidden relationships in real time using a point-and-click interface. Stanley recommends that first-time users start with Knowledge Explorer, which runs on laptop computers.
The company’s newest software component is the Applied Semantic Knowledgebase (ASK), which takes output obtained from earlier components of Sentient Suite, integrates the data, and generates patterns that detect, for example, drug toxicity. “We can help customers create profiles for different types of toxicity for different compounds.”
ASK provides an advantage in areas such as biomarker-based predictive biology and personalized medicine, profile creation and validation, compound efficiency and promiscuity screening, toxicity profiling and detection, disease signature detection, predictive clinical trials pre-screening, and patient stratification.
ASK starts with predictive network models identified using Sentient for data access and integration. Next, semantic SPARQL query technology is applied to build complex searches across multiple information sets. The SPARQL technology detects patterns within and between different types of data and relationships, even if the initial datasets were not formally related. These patterns or models are then placed into ASK, where researchers can query the data or apply the search patterns to new datasets for deeper investigation and analysis or decision making.
A key focus at IO Informatics is application of the semantics technology to healthcare and life science projects, such as personalized medicine, to identify biomarkers of early toxicity and efficacy.
“We believe we have the easiest to use and most accessible semantic integration platform. In addition to our transformative framework, we also have the ability to rapidly integrate and manage complex life science and healthcare formats, including chemical structures, DICOM images, and next-gen sequences, directly into the semantically integrated environment,” says Stanley.