%0 Journal Article
%T pISTil: a pipeline for yeast two-hybrid Interaction Sequence Tags identification and analysis
%A Johann Pellet
%A Laur¨¨ne Meyniel
%A Pierre-Olivier Vidalain
%A Beno£¿t de Chassey
%A Lionel Tafforeau
%A Vincent Lotteau
%A Chantal Rabourdin-Combe
%A Vincent Navratil
%J BMC Research Notes
%D 2009
%I BioMed Central
%R 10.1186/1756-0500-2-220
%X We develop pISTil, a bioinformatics pipeline combined with a user-friendly web-interface: (i) to establish a standardised system to analyse and to annotate ISTs generated by two-hybrid technologies with high performance and flexibility and (ii) to provide high-quality protein-protein interaction datasets for systems-level approach. This pipeline has been validated on a large dataset comprising more than 11.000 ISTs. As a case study, a detailed analysis of ISTs obtained from yeast two-hybrid screens of Hepatitis C Virus proteins against human cDNA libraries is also provided.We have developed pISTil, an open source pipeline made of a collection of several applications governed by a Perl script. The pISTil pipeline is intended to laboratories, with IT-expertise in system administration, scripting and database management, willing to automatically process large amount of ISTs data for accurate reconstruction of protein interaction networks in a systems biology perspective. pISTil is publicly available for download at http://sourceforge.net/projects/pistil webcite.Systems biology focuses, in part, on exhaustive and accurate reconstruction of molecular interaction networks, which support cellular machinery, i.e interactomes, under physiological or pathological conditions.Molecular interactions data related to human and model organisms are currently being integrated in generalist databases, such as INTACT [1], MINT [2] or STRING [3]. Some other databases are more specialised, as for instance VirHostNet, a knowledgebase devoted to virus-host interactions that allows analysis and visualisation of infection at the systems level [4]. One of the main sources of protein-protein interactions deposited in these public databases is generated by yeast two-hybrid (Y2H) technology. Indeed, Y2H allows high-throughput screening of direct physical protein-protein interactions at a proteome scale, but requires the sequencing of hundreds to thousands of cellular preys per experiment. These
%U http://www.biomedcentral.com/1756-0500/2/220