Cmero M1,2, McKay M1, Milton M1, Rajasekhar P1,2, Whitehead L1,2, Watson E1,2, Lam P1, Iskander J1,2, LIesse-Labat M1,2, Rogers K1,2, Papenfuss A1,2
1Walter and Eliza Hall Institute of Medical Research, Parkville, Australia, 2Department of Medical Biology, The University of Melbourne, Parkville, Australia
Biography:
Marek is a Senior Research Officer based in the Genomics lab at the Walter and Eliza Hall Institute of Medical Research. With a background in computer science and bioinformatics, he completed his PhD in cancer genomics at the University of Melbourne. Currently, he leads pipeline development for the Spatial Omics Data Analytics (SODA) Hub at WEHI, and the Advanced Genomics Facility. Marek's focus is on developing robust and scalable software to drive biological discoveries.
Abstract:
Spatial omics technologies are growing in use in biomedical research, allowing researchers to investigate tissues in unprecedented detail. These technologies present unique challenges in data management, processing and analysis due to their size, heterogeneous data formats, and a rapidly evolving tool ecosystem. To address this, we have established the WEHI Spatial Omics Data Analytics (SODA) Hub to provide a streamlined environment for spatial analysis, from data ingestion to analysis pipelines.
Beyond ingestion, a critical bottleneck remains in processing and analysing large, complex datasets that these instruments produce. Here we describe the open-source, scalable analysis pipelines we have developed to address this challenge. To make pipelines user friendly, robust and scalable, we utilise the Nextflow workflow system via Seqera Platform, which provides a user-friendly, cloud-hosted interface for workflow submission and monitoring, whilst all computation runs on local HPC infrastructure.
We present two pipelines, WEHI-SODA-Hub/spatialvpt for the MERSCOPE instrument and WEHI-SODA-Hub/sp_segment for spatial proteomics to perform cell segmentation, which involves a complex, multi-step toolchain. We have developed custom tools to facilitate parallelised cell measurements for large images. Downstream, we developed a cell phenotyping workflow, via a custom plugin for QuPath. This provides researchers with a familiar, GUI-based environment for their analysis. Users can access QuPath via Open OnDemand, and load images directly from OMERO.
These workflows illustrate how the SODA Hub is building tools to accelerate spatial research by reducing the time needed to manage, import, process, and analyse data. These open-source tools are available to the broader research community at github.com/WEHI-SODA-Hub.