|
Welcome to BioISA, the biological sciences division of the Institute for Security & Analysis.
We are very excited to be able to provide exclusive access to a powerful new biomedical information exploration and analysis platform. We have adapted and altered cutting-edge software previously reserved for hedge fund managers and intelligence community officials to meet the needs of the biomedical community. The BioISA analytical platform is based upon a novel approach to organizing, exploring, and analyzing massive, disparate datasets. By allowing innate human intelligence to interact seamlessly with powerful, intuitive technology, BioISA enables unprecedented analytic capabilities across both structured and unstructured data. The platform has been deployed in the financial sector, where high-performance hedge funds use it to perform sophisticated modeling and analysis on countless indicators, as well as in the intelligence community, which uses a different incarnation of the software to track patterns and anomalies in individual and group behaviour in an effort to identify and prevent criminal activity. BioISA's ability to incorporate and resolve data from multiple disparate sources makes it an ideal platform for exploring the vast amounts of biological information available to researchers. Rather than spend hours tracking down and manually synthesizing all the available data on a particular gene, researchers now have the entire NCBI human organism, WikiPathways, CTD, UniProt, KEGG PAthway, and PharmGKB databases, along with millions of PubMed abstracts and a variety of other data, at their fingertips. Within the fully interactive, explorable environment of BioISA, each gene, protein, disease, drug, and abstract becomes a manipulatable object with a host of structured properties that can be searched upon and analyzed, allowing for deep, multi-level investigation. Research abstracts can not only be browsed but relevant text within them can be tagged and resolved to existing objects, bridging the gap between structured and unstructured data. Explore interactions; browse abstracts; discover distant relationships. The project begins with an exploration of the human organism. |
Check out BioISA's latest initiative, BodyHacks:
|
| Path Search | Browsing | |
| It is well-known that a large percentage of diabetics die from heart disease, but suppose you have been tasked with investigating why this relationship exists - in other words, how are diabetes and heart failure connected? Manually determining all potential chains of genetic interactions that connect diabetes to heart failure through several degrees of separation would be a painstaking, time-intensive process. Such as process is made trivial with Palantir Biology's Path Search capability. | For more exploratory oriented tasks, BioISA offers a broad range of powerful tools. Each object in BioISA has a set of properties associated with it that are populated with structured data taken from one of the datasets being utilized. The SearchAround feature of BioISA allows the researcher to use these properties to link objects together, creating a detailed representation of a particular object's relationships. SearchArounds can be tightly controlled through the use of linking and target parameters, allowing both very specific or very broad searches to be performed. | |
 |  | |
| With just a few clicks, we can lay bare all potential relationships connecting diabetes (on the left) to heart failure (on the right). Above, see the genes connecting heart failure to diabetes. | Above, we see a representation of the all the relationships of the disease osteoarthritis, uncovered through a depth 3 SearchAround with just a few clicks. | |
 |  | |
| Above, the results of a Path Search between Multiple Sclerosis and the drug Atorvastatin. | We can now dig deeper into individual clusters and objects. On closer examination of the graph, we see that the gene tp53 is indirectly related to osteoarthritis. Selecting the tp53 object brings up a detailed breakdown of all the information associated with it. | |
| Supplementing BioISA's powerful analytic features are deep collaborative capabilities. As research is often performed with colleagues within a lab, research unit, or department, it is important for researchers to be able to easily share their findings with others. BioISA lets you selectively choose which findings to publish, and other's findings selectively accepted, allowing granular control.
Moreover, every action performed is automatically saved, allowing for the sharing of entire investigations as well as the ability to return to any point in an investigation, allowing for the branching of different methods | ||