Florian Leitner completed his PhD at the CNIO in Spain with Alfonso Valencia in 2012, specializing in text mining. He obtained his master's degree in computational biology in 2006, with Frank Eisenhaber at the Institute for Molecular Pathology (IMP) in Vienna, Austria. His current core interest is developing a robust, high-throughput text mining pipeline for extracting biologically relevant relationships from full-text (as opposed to abstracts), with a very particular focus on transcriptional regulation events. He is also involved in the organization of BioCreative (www.biocreative.org), a community platform for text mining in biology.

The title of his talk: "Reconstructing cis-regulatory transcription networks from literature"

Cis-regulatory transcription events are at the core of intracellular gene regulation networks. These networks are being traced with many different experimental methods, while knowledge of the underlying transcription events is essential to connect (co-) factors to their associated target genes in in silico interaction models. Frequently, high throughput methods (e.g., ChIP-on-chip and ChIP-seq) are used to elucidate these relationships. However, these methods commonly lack precision and/or coverage and thus may fail to capture already known regulation events or produce false positives. Database efforts to manually curate the existing small-scale data published over the last few decades are not capable of processing the large numbers of already published data, and are mostly only available as proprietary resources. Therefore, here we present a pattern-based text mining approach to extract the cis-regulatory transcriptional regulation events (TREs) using finite state automata. Compared to existing work, this approach attempts to extract the normalized (i.e., sequence database-mapped), cis-regulatory events, thereby aiming well beyond the capabilities of current approaches.