Using the Elements of Style in Workflow creation approach as taught in the https://nih-nichd.github.io, the team will work together to containerize the processing steps and stitch them together in a workflow language. On Cavatica, the workflow language best supported is Common Workflow Language (CWL).
 |
 |
 |
MYC is an oncoprotein and often implies worse outcomes, however, it also seems to have a role in cardiovascular disease. Using open data from both the INCLUDE Data Hub and the Kids First (KF) Data Resource Portal available through the Cavatica Platform, this Hackathon will:
-
Explore MYC expression and its segregating effects at the intersection of A) congenital heart defects and cancer from KF & B) & Down Syndrome (Include)
-
Build platform agnostic workflows and analysis notebooks using the elements of style method (https://nih-nichd.github.io) that will use both WGCNA and limma to probe the dual role of this gene.
-
Result in a tutorial of how to proceed from both https://portal.includedcc.org/ and https://portal.kidsfirstdrc.org to https://cavatica.sbgenomics.com/ as well as produce results in notebooks that could be used as figures in a paper promoting open transparent, reproducible results.
- Make two matrices of gene x sample in Cavatica Portal -- in our case the Kids First RNA-seq workflow (pre-made by CHOP).
- Combine expression matrices for birth defects and cancer.
- Normalize gene expression with deseq2 (not using for differential expression, only for quantile normalisation).
- Separate the samples between high and low MYC expression using Gene Median Splitter.
- Classify all subjects by phenotype.
- Run WGNCA analysis after binning based on High MYC and Low MYC expression.
- Create a template jupyter lab notebook https://bioinformaticsworkbook.org/tutorials/wgcna.html#gsc.tab=0).
- Look at congenital and cardiac outcomes that might be due to drug interactions. *Related Repo for MYC related pediatric drug saftey profiles
- What are some outcome differences between high and low MYC expression; different heart morphologies, birth defects based on HPO terms.
- Run annotation analysis such as GO and biological pathways, maybe Reactome.
Workflow Diagram:
Analysis plan - Weighted Gene Co-expression Network Analysis (WGCNA) See Publication exists as an R program -> convert to a notebook -> break out the DESeq from the WGCNA
Treatment in the WGCNA sense is the condition of the sample:
- Trisomy and Disomy samples
- High Myc and Low Myc expression.
- Birth defect (cardiac) and Blood Cancers
- Birth Defect (cardiac) and Solid Tumors