Presentation Abstract

Program#/Poster#: 328.09/A9
Presentation Title: Hypothesis generation and reconstruction of the connectome and cognome from the literature
Location: Hall A-C
Presentation time: Monday, Nov 14, 2011, 8:00 AM - 9:00 AM
Authors: *B. VOYTEK1, J. B. VOYTEK2;
1Univ. California, Berkeley, Berkeley, CA; 2Sch. of Information, Univ. of California, Berkeley, Berkeley, CA
Abstract: Modern scientific research stands on the shoulders of countless giants. PubMed alone contains more than 18 million peer-reviewed articles. In order to synthesize the breadth of neuroscientific research, we populated a database of associations between neuroscientific concepts built from over 3.5 million scientific abstracts. From the literature, we reconstructed a systems-level connectome as well as_for the first time_associations between brain structure, cognitive functions, disease, and gene expression data from the Allen Brain Atlas. We introduce an algorithm for semi-automated hypothesis generation based upon the statistical properties of this “cognome” to identify possibly novel research associations. Here we take an important step toward incorporating data-mining algorithms into the scientific method in a manner that is generalizable to any scientific field.
The scientific method begins with a hypothesis about our world that can be experimentally tested. Hypothesis formation is iterative, building off of prior scientific knowledge. Thus, before one can even begin to test a hypothesis, one must have a thorough understanding of previous research to ensure that the path of inquiry is founded upon a stable base of established facts. But how can a researcher perform a thorough, unbiased literature review when over one million scientific articles are published annually? The rate of scientific discovery has quickly outpaced our ability to integrate even the newest findings in an unbiased, principled fashion.
In this paper we show that the literature contains a vast amount of connected facts that, by definition, recapitulate known neuroscientific relationships. Neuroanatomical, behavioral, and disease associations can be rapidly quantified and visualized to speed research and education or to find understudied research paths. Rather than allowing our limited ability to thoroughly review the entire scientific literature bias our hypotheses, we can algorithmically integrate across millions of scientific research papers in a principled fashion.
Disclosures:  B. Voytek: None. J.B. Voytek: None.
Keyword(s): CONNECTION
Support: NINDS Grant NS21135-22S1
Society for Neuroscience Neuroscience Scholars Program
[Authors]. [Abstract Title]. Program No. XXX.XX. 2011 Neuroscience Meeting Planner. Washington, DC: Society for Neuroscience, 2011. Online.

2011 Copyright by the Society for Neuroscience all rights reserved. Permission to republish any abstract or part of any abstract in any form must be obtained in writing by SfN office prior to publication.