The Bioinformatics Platform in Psychiatry and Neuroscience (BIPN) is affiliated with INSERM U1266 – Paris Institute of Psychiatry and Neuroscience (IPNP). Its purpose is to provide bioinformatics, methodological, and analytical support to research teams developing projects in psychiatry, fundamental and clinical neuroscience, and the field of neurodevelopmental disorders. The platform provides cross-disciplinary support for projects involving high-throughput biological data, whether genetic, transcriptomic, epigenomic, proteomic, or multi-omic. It supports teams from the project design phase, helping to formulate scientific questions and define appropriate analytical strategies, through to data processing, statistical analysis, and biological interpretation of results, with a focus on scientific quality, reproducibility, and sustainability of methods.
On the technical side, the platform is managed by Benjamin DEMAILLE, Technical Manager at BIPN. On the scientific side, coordination is provided by Nicolas Ramoz, Internal Scientific Manager at IPNP, in collaboration with Anton Iftimovici, Associate Scientific Manager at GHU-Paris.
The main mission of the BIPN platform is to formalize and structure the analytical needs of research teams, working closely with biologists, clinicians, and project leaders. This mission includes assisting in the selection of the most appropriate omics technologies, defining analysis plans, anticipating constraints related to data volumes, cohort heterogeneity, batch effects, and technical biases, as well as implementing robust and statistically relevant analytical strategies. The platform also processes, analyzes, and integrates complex biological data, ensuring the quality, traceability, and reproducibility of analyses. An essential complementary mission involves maintaining, developing, and sharing bioinformatics tools, pipelines, and best practices in order to promote the development of IPNP teams' skills and ensure the sustainability of the analyses produced.
The platform's activities cover a broad spectrum of bioinformatics applied to the life sciences. In human genomics, BIPN supports projects involving genome-wide association studies, polygenic risk score calculations, and whole genome sequencing data analysis. These analyses include quality control of genotypic or sequencing data, population structure management, variant calling, filtering, and annotation, as well as biological interpretation of results in psychiatric and neurological contexts. In transcriptomics, the platform supports bulk and single-cell RNA-seq data analysis, from data preprocessing and quantification to expression signature identification, cell annotation, and integration with other omics levels. The platform also offers support for the analysis of epigenomic and methylomic data, including quality control, normalization, correction of technical effects, and differential analyses, as well as for the analysis of proteomic data and their integration into multi-omic approaches.
Data engineering and reproducible analyses
A central focus of the platform's activity concerns data engineering and the implementation of reproducible analyses. BIPN ensures the structuring and urbanization of data, the design of architectures adapted to heterogeneous and large datasets, as well as the development of standardized, documented, and versioned analysis pipelines. Particular attention is paid to metadata management, format consistency, and processing traceability in order to facilitate the reuse of data and results in future projects. This approach guarantees the methodological robustness of the analyses and their compliance with current requirements for reproducible science and best practices in bioinformatics.
Training and support
The platform places great importance on training and supporting users. It offers ongoing methodological support to teams, including assistance with analysis design, the provision of reproducible scripts and workflows, and the drafting of documentation and tutorials. The platform also participates in training researchers, engineers, and students in bioinformatics tools and methods, with a view to transferring skills and strengthening the autonomy of research teams. These actions contribute to disseminating best practices in data analysis, data management, and reproducibility within the IPNP.
Collaborations
The BIPN platform works closely with several IPNP research teams, including the Krebs, Gorwood/Ramoz, and Oppenheim teams, as well as with GHU-Paris (website: ghu-paris.fr) on projects involving clinical and biological data. It is also involved in national and external collaborations, in particular with PEPR-PROPSY, CNRGH, and the Imagine Institute, as well as with industrial partners. These collaborations promote the development of interdisciplinary projects and the integration of the platform into large-scale research networks.
The BIPN platform has technical resources dedicated to the analysis of large-scale biological data. It is supported by a secure storage infrastructure of approximately 200 terabytes, enabling the archiving and management of large datasets such as whole genome sequencing data, single-cell transcriptomics, and multi-omics projects. Computational analyses are performed on a Mac Studio M3 Ultra high-performance computing station equipped with 32 CPU cores and 512 GB of RAM, supplemented by a MacBook Pro M4 Max personal workstation with 128 GB of RAM, used for development, pipeline testing, and targeted analyses. These resources enable the platform to meet the varied computational needs of the projects it supports while promoting the reproducibility of analyses.
Environments and tools
The analyses performed within the platform are based on Linux environments and the use of standard programming languages in bioinformatics, including R, Python, and SQL. The platform uses and maintains tools that are widely recognized by the scientific community, such as CellRanger for single-cell transcriptomics, Plink and PRSice2 for human genetics and polygenic risk scores, Minfi for DNA methylation analysis, Salmon for rapid and robust quantification of bulk transcriptomics data, Bismark for bisulfite sequencing data processing, and ENSEMBL-VEP for functional annotation of variants. These tools are integrated into reproducible and documented workflows, tailored to the specific needs of projects conducted within the IPNP and compliant with community standards.
The platform is primarily open to IPNP teams, as well as academic and clinical partners within the framework of established collaborations. Projects are accepted after a discussion phase to define scientific objectives, available data, analysis methods, and expected deliverables. Users agree to comply with the platform's operating rules, particularly with regard to data management, confidentiality, and security, in accordance with Inserm and IPNP institutional recommendations.
