Leveraging the Cloud for Splicing Discovery

项目总结:

PROJECT SUMMARY: Alternative splicing is among the most important contributors of proteomic diversity in higher order eukaryotes. When disrupted in cancer, mis-splicing results in unique mRNA isoforms not observed in healthy cells. Such cancer-specific splice isoforms represent an untapped reservoir of potential neoantigens for targeted cancer vaccines and immunotherapies. As a central component of our funded NCI R01 for “Unbiased identification of spliceosome vulnerabilities across cancer”, we have been extending and leveraging a comprehensive splicing analysis pipeline to define splicing vulnerabilities across human cancers and healthy tissues. The associated bioinformatics tools that are built-upon through this effort are designed to identify both known and novel cancer subtypes, nominate key regulatory splicing factors, infer functional splice-isoform impacts and discover cancer- specific neoantigens that can be exploited by emerging immunotherapies or vaccines. The bioinformatics tools to yield these discoveries largely consistent of distinct components of the large AltAnalyze open-source project, begun in 2008. While AltAnalyze or its algorithms have been cited in over 400 published research studies, re- applying this workflow at the scale of TCGA, GTEx and other large RNA-Seq compendiums have historically required significant computational resources and time to download and re-process hundreds of terabytes of data in a secure and compliant manner. New emerging cloud-based solutions have the potential to mitigate these technical challenges through streamlined compute of thousands of pre-processed samples at a low cost. To address these challenges, we propose to: Aim 1: Streamline and optimize AltAnalyze for the cloud. In this aim, we will decouple and optimize the primary splicing analysis components of AltAnalyze to enable streamlined supervised and unsupervised analysis of cancer transcriptomes. AltAnalyze will be packaged as a CWL pipeline, containerized with Docker and deposited in DockStore, enabling fast and comprehensive analyses of splicing in the cloud. Aim 2: Integrate AltAnalyze.cloud in Terra.bio. To enable direct analyses of controlled-access sequence- level files in TCGA, TARGET, GTEx and other major human RNA-sequencing datasets, we will 1) translate our CWL workflows to the Workflow Description Language (WDL) and 2) establish a Terra workflow for integrated splicing analysis using AltAnalyze.cloud. AltAnalyze.cloud will be able to be run through the Terra web interface for analysis, progress tracking, provenance and sharing of results. These features will enable streamlined re-use of user and controlled NIH deposited datasets in the cloud.