Implementing a transcription factor interaction prediction system using the genometric query language

Novel technologies and growing interest have resulted in a large increase in the amount of data available for genomics and transcriptomics studies, both in terms of volume and contents. Biology is relying more and more on computational methods to process, investigate, and extract knowledge from this huge amount of data. In this work, we present the TICA web server (available at http://www.gmql.eu/tica/), a fast and compact tool developed to support data-driven knowledge discovery in the realm of transcription factor interaction prediction. TICA leverages both the GenoMetric Query Language, a novel query tool (based on the Apache Hadoop and Spark technologies) specialized in the integration and management of heterogeneous, large genomic datasets, and a statistical method for robust detection of co-locations across interval-based data, in order to infer physically interacting transcription factors. Notably, TICA allows investigators to upload and analyze their own ChIP-seq experiments datasets, comparing them both against ENCODE data or between themselves, achieving computation time which increases linearly with respect to dataset size and density. Using ENCODE data from three well-studied cell lines as reference, we show that TICA predictions are supported by existing biological knowledge, making the web server a reliable and efficient tool for interaction screening and data-driven hypothesis generation.